kernel - SWAP CACHE part 3/many - Rearrange VM pagerops

[dragonfly.git] / sys / vm / vnode_pager.c

/*
 * Copyright (c) 1990 University of Utah.
 * Copyright (c) 1991 The Regents of the University of California.
 * All rights reserved.
 * Copyright (c) 1993, 1994 John S. Dyson
 * Copyright (c) 1995, David Greenman
 *
 * This code is derived from software contributed to Berkeley by
 * the Systems Programming Group of the University of Utah Computer
 * Science Department.
 *
 * 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.
 *
 *      from: @(#)vnode_pager.c 7.5 (Berkeley) 4/20/91
 * $FreeBSD: src/sys/vm/vnode_pager.c,v 1.116.2.7 2002/12/31 09:34:51 dillon Exp $
 * $DragonFly: src/sys/vm/vnode_pager.c,v 1.43 2008/06/19 23:27:39 dillon Exp $
 */

/*
 * Page to/from files (vnodes).
 */

/*
 * TODO:
 *      Implement VOP_GETPAGES/PUTPAGES interface for filesystems. Will
 *      greatly re-simplify the vnode_pager.
 */

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/proc.h>
#include <sys/vnode.h>
#include <sys/mount.h>
#include <sys/buf.h>
#include <sys/vmmeter.h>
#include <sys/conf.h>
#include <sys/sfbuf.h>

#include <vm/vm.h>
#include <vm/vm_object.h>
#include <vm/vm_page.h>
#include <vm/vm_pager.h>
#include <vm/vm_map.h>
#include <vm/vnode_pager.h>
#include <vm/vm_extern.h>

#include <sys/thread2.h>
#include <vm/vm_page2.h>

static void vnode_pager_dealloc (vm_object_t);
static int vnode_pager_getpage (vm_object_t, vm_page_t *, int);
static void vnode_pager_putpages (vm_object_t, vm_page_t *, int, boolean_t, int *);
static boolean_t vnode_pager_haspage (vm_object_t, vm_pindex_t);

struct pagerops vnodepagerops = {
        vnode_pager_alloc,
        vnode_pager_dealloc,
        vnode_pager_getpage,
        vnode_pager_putpages,
        vnode_pager_haspage
};

static struct krate vbadrate = { 1 };
static struct krate vresrate = { 1 };

int vnode_pbuf_freecnt = -1;    /* start out unlimited */

/*
 * Allocate (or lookup) pager for a vnode.
 * Handle is a vnode pointer.
 */
vm_object_t
vnode_pager_alloc(void *handle, off_t size, vm_prot_t prot, off_t offset)
{
        vm_object_t object;
        struct vnode *vp;

        /*
         * Pageout to vnode, no can do yet.
         */
        if (handle == NULL)
                return (NULL);

        /*
         * XXX hack - This initialization should be put somewhere else.
         */
        if (vnode_pbuf_freecnt < 0) {
            vnode_pbuf_freecnt = nswbuf / 2 + 1;
        }

        vp = (struct vnode *) handle;

        /*
         * Prevent race condition when allocating the object. This
         * can happen with NFS vnodes since the nfsnode isn't locked.
         */
        while (vp->v_flag & VOLOCK) {
                vsetflags(vp, VOWANT);
                tsleep(vp, 0, "vnpobj", 0);
        }
        vsetflags(vp, VOLOCK);

        /*
         * If the object is being terminated, wait for it to
         * go away.
         */
        while (((object = vp->v_object) != NULL) &&
                (object->flags & OBJ_DEAD)) {
                vm_object_dead_sleep(object, "vadead");
        }

        if (vp->v_sysref.refcnt <= 0)
                panic("vnode_pager_alloc: no vnode reference");

        if (object == NULL) {
                /*
                 * And an object of the appropriate size
                 */
                object = vm_object_allocate(OBJT_VNODE,
                                            OFF_TO_IDX(round_page(size)));
                object->flags = 0;
                object->handle = handle;
                vp->v_object = object;
                vp->v_filesize = size;
        } else {
                object->ref_count++;
                if (vp->v_filesize != size) {
                        kprintf("vnode_pager_alloc: Warning, filesize "
                                "mismatch %lld/%lld\n",
                                (long long)vp->v_filesize,
                                (long long)size);
                }
        }
        vref(vp);

        vclrflags(vp, VOLOCK);
        if (vp->v_flag & VOWANT) {
                vclrflags(vp, VOWANT);
                wakeup(vp);
        }
        return (object);
}

static void
vnode_pager_dealloc(vm_object_t object)
{
        struct vnode *vp = object->handle;

        if (vp == NULL)
                panic("vnode_pager_dealloc: pager already dealloced");

        vm_object_pip_wait(object, "vnpdea");

        object->handle = NULL;
        object->type = OBJT_DEAD;
        vp->v_object = NULL;
        vp->v_filesize = NOOFFSET;
        vclrflags(vp, VTEXT | VOBJBUF);
}

/*
 * Return whether the vnode pager has the requested page.  Return the
 * number of disk-contiguous pages before and after the requested page,
 * not including the requested page.
 */
static boolean_t
vnode_pager_haspage(vm_object_t object, vm_pindex_t pindex)
{
        struct vnode *vp = object->handle;
        off_t loffset;
        off_t doffset;
        int voff;
        int bsize;
        int error;

        /*
         * If no vp or vp is doomed or marked transparent to VM, we do not
         * have the page.
         */
        if ((vp == NULL) || (vp->v_flag & VRECLAIMED))
                return FALSE;

        /*
         * If filesystem no longer mounted or offset beyond end of file we do
         * not have the page.
         */
        loffset = IDX_TO_OFF(pindex);

        if (vp->v_mount == NULL || loffset >= vp->v_filesize)
                return FALSE;

        bsize = vp->v_mount->mnt_stat.f_iosize;
        voff = loffset % bsize;

        /*
         * XXX
         *
         * BMAP returns byte counts before and after, where after
         * is inclusive of the base page.  haspage must return page
         * counts before and after where after does not include the
         * base page.
         *
         * BMAP is allowed to return a *after of 0 for backwards
         * compatibility.  The base page is still considered valid if
         * no error is returned.
         */
        error = VOP_BMAP(vp, loffset - voff, &doffset, NULL, NULL, 0);
        if (error)
                return TRUE;
        if (doffset == NOOFFSET)
                return FALSE;
        return TRUE;
}

/*
 * Lets the VM system know about a change in size for a file.
 * We adjust our own internal size and flush any cached pages in
 * the associated object that are affected by the size change.
 *
 * NOTE: This routine may be invoked as a result of a pager put
 * operation (possibly at object termination time), so we must be careful.
 *
 * NOTE: vp->v_filesize is initialized to NOOFFSET (-1), be sure that
 * we do not blow up on the case.  nsize will always be >= 0, however.
 */
void
vnode_pager_setsize(struct vnode *vp, vm_ooffset_t nsize)
{
        vm_pindex_t nobjsize;
        vm_pindex_t oobjsize;
        vm_object_t object = vp->v_object;

        if (object == NULL)
                return;

        /*
         * Hasn't changed size
         */
        if (nsize == vp->v_filesize)
                return;

        /*
         * Has changed size.  Adjust the VM object's size and v_filesize
         * before we start scanning pages to prevent new pages from being
         * allocated during the scan.
         */
        nobjsize = OFF_TO_IDX(nsize + PAGE_MASK);
        oobjsize = object->size;
        object->size = nobjsize;

        /*
         * File has shrunk. Toss any cached pages beyond the new EOF.
         */
        if (nsize < vp->v_filesize) {
                vp->v_filesize = nsize;
                if (nobjsize < oobjsize) {
                        vm_object_page_remove(object, nobjsize, oobjsize,
                                              FALSE);
                }
                /*
                 * This gets rid of garbage at the end of a page that is now
                 * only partially backed by the vnode.  Since we are setting
                 * the entire page valid & clean after we are done we have
                 * to be sure that the portion of the page within the file
                 * bounds is already valid.  If it isn't then making it
                 * valid would create a corrupt block.
                 */
                if (nsize & PAGE_MASK) {
                        vm_offset_t kva;
                        vm_page_t m;

                        do {
                                m = vm_page_lookup(object, OFF_TO_IDX(nsize));
                        } while (m && vm_page_sleep_busy(m, TRUE, "vsetsz"));

                        if (m && m->valid) {
                                int base = (int)nsize & PAGE_MASK;
                                int size = PAGE_SIZE - base;
                                struct sf_buf *sf;

                                /*
                                 * Clear out partial-page garbage in case
                                 * the page has been mapped.
                                 *
                                 * This is byte aligned.
                                 */
                                vm_page_busy(m);
                                sf = sf_buf_alloc(m, SFB_CPUPRIVATE);
                                kva = sf_buf_kva(sf);
                                bzero((caddr_t)kva + base, size);
                                sf_buf_free(sf);

                                /*
                                 * XXX work around SMP data integrity race
                                 * by unmapping the page from user processes.
                                 * The garbage we just cleared may be mapped
                                 * to a user process running on another cpu
                                 * and this code is not running through normal
                                 * I/O channels which handle SMP issues for
                                 * us, so unmap page to synchronize all cpus.
                                 *
                                 * XXX should vm_pager_unmap_page() have
                                 * dealt with this?
                                 */
                                vm_page_protect(m, VM_PROT_NONE);

                                /*
                                 * Clear out partial-page dirty bits.  This
                                 * has the side effect of setting the valid
                                 * bits, but that is ok.  There are a bunch
                                 * of places in the VM system where we expected
                                 * m->dirty == VM_PAGE_BITS_ALL.  The file EOF
                                 * case is one of them.  If the page is still
                                 * partially dirty, make it fully dirty.
                                 *
                                 * NOTE: We do not clear out the valid
                                 * bits.  This would prevent bogus_page
                                 * replacement from working properly.
                                 *
                                 * NOTE: We do not want to clear the dirty
                                 * bit for a partial DEV_BSIZE'd truncation!
                                 * This is DEV_BSIZE aligned!
                                 */
                                vm_page_clear_dirty_beg_nonincl(m, base, size);
                                if (m->dirty != 0)
                                        m->dirty = VM_PAGE_BITS_ALL;
                                vm_page_wakeup(m);
                        }
                }
        } else {
                vp->v_filesize = nsize;
        }
}

/*
 * Release a page busied for a getpages operation.  The page may have become
 * wired (typically due to being used by the buffer cache) or otherwise been
 * soft-busied and cannot be freed in that case.  A held page can still be
 * freed.
 */
void
vnode_pager_freepage(vm_page_t m)
{
        if (m->busy || m->wire_count) {
                vm_page_activate(m);
                vm_page_wakeup(m);
        } else {
                vm_page_free(m);
        }
}

/*
 * EOPNOTSUPP is no longer legal.  For local media VFS's that do not
 * implement their own VOP_GETPAGES, their VOP_GETPAGES should call to
 * vnode_pager_generic_getpages() to implement the previous behaviour.
 *
 * All other FS's should use the bypass to get to the local media
 * backing vp's VOP_GETPAGES.
 */
static int
vnode_pager_getpage(vm_object_t object, vm_page_t *mpp, int seqaccess)
{
        int rtval;
        struct vnode *vp;

        vp = object->handle;
        rtval = VOP_GETPAGES(vp, mpp, PAGE_SIZE, 0, 0, seqaccess);
        if (rtval == EOPNOTSUPP)
                panic("vnode_pager: vfs's must implement vop_getpages\n");
        return rtval;
}

/*
 * This is now called from local media FS's to operate against their
 * own vnodes if they fail to implement VOP_GETPAGES.
 *
 * With all the caching local media devices do these days there is really
 * very little point to attempting to restrict the I/O size to contiguous
 * blocks on-disk, especially if our caller thinks we need all the specified
 * pages.  Just construct and issue a READ.
 */
int
vnode_pager_generic_getpages(struct vnode *vp, vm_page_t *mpp, int bytecount,
                             int reqpage, int seqaccess)
{
        struct iovec aiov;
        struct uio auio;
        off_t foff;
        int error;
        int count;
        int i;
        int ioflags;

        /*
         * Do not do anything if the vnode is bad.
         */
        if (vp->v_mount == NULL)
                return VM_PAGER_BAD;

        /*
         * Calculate the number of pages.  Since we are paging in whole
         * pages, adjust bytecount to be an integral multiple of the page
         * size.  It will be clipped to the file EOF later on.
         */
        bytecount = round_page(bytecount);
        count = bytecount / PAGE_SIZE;

        /*
         * We could check m[reqpage]->valid here and shortcut the operation,
         * but doing so breaks read-ahead.  Instead assume that the VM
         * system has already done at least the check, don't worry about
         * any races, and issue the VOP_READ to allow read-ahead to function.
         *
         * This keeps the pipeline full for I/O bound sequentially scanned
         * mmap()'s
         */
        /* don't shortcut */

        /*
         * Discard pages past the file EOF.  If the requested page is past
         * the file EOF we just leave its valid bits set to 0, the caller
         * expects to maintain ownership of the requested page.  If the
         * entire range is past file EOF discard everything and generate
         * a pagein error.
         */
        foff = IDX_TO_OFF(mpp[0]->pindex);
        if (foff >= vp->v_filesize) {
                for (i = 0; i < count; i++) {
                        if (i != reqpage)
                                vnode_pager_freepage(mpp[i]);
                }
                return VM_PAGER_ERROR;
        }

        if (foff + bytecount > vp->v_filesize) {
                bytecount = vp->v_filesize - foff;
                i = round_page(bytecount) / PAGE_SIZE;
                while (count > i) {
                        --count;
                        if (count != reqpage)
                                vnode_pager_freepage(mpp[count]);
                }
        }

        /*
         * The size of the transfer is bytecount.  bytecount will be an
         * integral multiple of the page size unless it has been clipped
         * to the file EOF.  The transfer cannot exceed the file EOF.
         *
         * When dealing with real devices we must round-up to the device
         * sector size.
         */
        if (vp->v_type == VBLK || vp->v_type == VCHR) {
                int secmask = vp->v_rdev->si_bsize_phys - 1;
                KASSERT(secmask < PAGE_SIZE, ("vnode_pager_generic_getpages: sector size %d too large\n", secmask + 1));
                bytecount = (bytecount + secmask) & ~secmask;
        }

        /*
         * Severe hack to avoid deadlocks with the buffer cache
         */
        for (i = 0; i < count; ++i) {
                vm_page_t mt = mpp[i];

                vm_page_io_start(mt);
                vm_page_wakeup(mt);
        }

        /*
         * Issue the I/O with some read-ahead if bytecount > PAGE_SIZE
         */
        ioflags = IO_VMIO;
        if (seqaccess)
                ioflags |= IO_SEQMAX << IO_SEQSHIFT;

        aiov.iov_base = NULL;
        aiov.iov_len = bytecount;
        auio.uio_iov = &aiov;
        auio.uio_iovcnt = 1;
        auio.uio_offset = foff;
        auio.uio_segflg = UIO_NOCOPY;
        auio.uio_rw = UIO_READ;
        auio.uio_resid = bytecount;
        auio.uio_td = NULL;
        mycpu->gd_cnt.v_vnodein++;
        mycpu->gd_cnt.v_vnodepgsin += count;

        error = VOP_READ(vp, &auio, ioflags, proc0.p_ucred);

        /*
         * Severe hack to avoid deadlocks with the buffer cache
         */
        for (i = 0; i < count; ++i) {
                vm_page_t mt = mpp[i];

                while (vm_page_sleep_busy(mt, FALSE, "getpgs"))
                        ;
                vm_page_busy(mt);
                vm_page_io_finish(mt);
        }

        /*
         * Calculate the actual number of bytes read and clean up the
         * page list.  
         */
        bytecount -= auio.uio_resid;

        for (i = 0; i < count; ++i) {
                vm_page_t mt = mpp[i];

                if (i != reqpage) {
                        if (error == 0 && mt->valid) {
                                if (mt->flags & PG_WANTED)
                                        vm_page_activate(mt);
                                else
                                        vm_page_deactivate(mt);
                                vm_page_wakeup(mt);
                        } else {
                                vnode_pager_freepage(mt);
                        }
                } else if (mt->valid == 0) {
                        if (error == 0) {
                                kprintf("page failed but no I/O error page %p object %p pindex %d\n", mt, mt->object, (int) mt->pindex);
                                /* whoops, something happened */
                                error = EINVAL;
                        }
                } else if (mt->valid != VM_PAGE_BITS_ALL) {
                        /*
                         * Zero-extend the requested page if necessary (if
                         * the filesystem is using a small block size).
                         */
                        vm_page_zero_invalid(mt, TRUE);
                }
        }
        if (error) {
                kprintf("vnode_pager_getpage: I/O read error\n");
        }
        return (error ? VM_PAGER_ERROR : VM_PAGER_OK);
}

/*
 * EOPNOTSUPP is no longer legal.  For local media VFS's that do not
 * implement their own VOP_PUTPAGES, their VOP_PUTPAGES should call to
 * vnode_pager_generic_putpages() to implement the previous behaviour.
 *
 * Caller has already cleared the pmap modified bits, if any.
 *
 * All other FS's should use the bypass to get to the local media
 * backing vp's VOP_PUTPAGES.
 */
static void
vnode_pager_putpages(vm_object_t object, vm_page_t *m, int count,
                     boolean_t sync, int *rtvals)
{
        int rtval;
        struct vnode *vp;
        int bytes = count * PAGE_SIZE;

        /*
         * Force synchronous operation if we are extremely low on memory
         * to prevent a low-memory deadlock.  VOP operations often need to
         * allocate more memory to initiate the I/O ( i.e. do a BMAP 
         * operation ).  The swapper handles the case by limiting the amount
         * of asynchronous I/O, but that sort of solution doesn't scale well
         * for the vnode pager without a lot of work.
         *
         * Also, the backing vnode's iodone routine may not wake the pageout
         * daemon up.  This should be probably be addressed XXX.
         */

        if ((vmstats.v_free_count + vmstats.v_cache_count) < vmstats.v_pageout_free_min)
                sync |= OBJPC_SYNC;

        /*
         * Call device-specific putpages function
         */
        vp = object->handle;
        rtval = VOP_PUTPAGES(vp, m, bytes, sync, rtvals, 0);
        if (rtval == EOPNOTSUPP) {
            kprintf("vnode_pager: *** WARNING *** stale FS putpages\n");
            rtval = vnode_pager_generic_putpages( vp, m, bytes, sync, rtvals);
        }
}


/*
 * This is now called from local media FS's to operate against their
 * own vnodes if they fail to implement VOP_PUTPAGES.
 *
 * This is typically called indirectly via the pageout daemon and
 * clustering has already typically occured, so in general we ask the
 * underlying filesystem to write the data out asynchronously rather
 * then delayed.
 */
int
vnode_pager_generic_putpages(struct vnode *vp, vm_page_t *m, int bytecount,
                             int flags, int *rtvals)
{
        int i;
        vm_object_t object;
        int maxsize, ncount, count;
        vm_ooffset_t poffset;
        struct uio auio;
        struct iovec aiov;
        int error;
        int ioflags;

        object = vp->v_object;
        count = bytecount / PAGE_SIZE;

        for (i = 0; i < count; i++)
                rtvals[i] = VM_PAGER_AGAIN;

        if ((int) m[0]->pindex < 0) {
                kprintf("vnode_pager_putpages: attempt to write meta-data!!! -- 0x%lx(%x)\n",
                        (long)m[0]->pindex, m[0]->dirty);
                rtvals[0] = VM_PAGER_BAD;
                return VM_PAGER_BAD;
        }

        maxsize = count * PAGE_SIZE;
        ncount = count;

        poffset = IDX_TO_OFF(m[0]->pindex);

        /*
         * If the page-aligned write is larger then the actual file we
         * have to invalidate pages occuring beyond the file EOF.
         *
         * If the file EOF resides in the middle of a page we still clear
         * all of that page's dirty bits later on.  If we didn't it would
         * endlessly re-write.
         *
         * We do not under any circumstances truncate the valid bits, as
         * this will screw up bogus page replacement.
         *
         * The caller has already read-protected the pages.  The VFS must
         * use the buffer cache to wrap the pages.  The pages might not
         * be immediately flushed by the buffer cache but once under its
         * control the pages themselves can wind up being marked clean
         * and their covering buffer cache buffer can be marked dirty.
         */
        if (poffset + maxsize > vp->v_filesize) {
                if (poffset < vp->v_filesize) {
                        maxsize = vp->v_filesize - poffset;
                        ncount = btoc(maxsize);
                } else {
                        maxsize = 0;
                        ncount = 0;
                }
                if (ncount < count) {
                        for (i = ncount; i < count; i++) {
                                rtvals[i] = VM_PAGER_BAD;
                        }
                }
        }

        /*
         * pageouts are already clustered, use IO_ASYNC to force a bawrite()
         * rather then a bdwrite() to prevent paging I/O from saturating
         * the buffer cache.  Dummy-up the sequential heuristic to cause
         * large ranges to cluster.  If neither IO_SYNC or IO_ASYNC is set,
         * the system decides how to cluster.
         */
        ioflags = IO_VMIO;
        if (flags & (VM_PAGER_PUT_SYNC | VM_PAGER_PUT_INVAL))
                ioflags |= IO_SYNC;
        else if ((flags & VM_PAGER_CLUSTER_OK) == 0)
                ioflags |= IO_ASYNC;
        ioflags |= (flags & VM_PAGER_PUT_INVAL) ? IO_INVAL: 0;
        ioflags |= IO_SEQMAX << IO_SEQSHIFT;

        aiov.iov_base = (caddr_t) 0;
        aiov.iov_len = maxsize;
        auio.uio_iov = &aiov;
        auio.uio_iovcnt = 1;
        auio.uio_offset = poffset;
        auio.uio_segflg = UIO_NOCOPY;
        auio.uio_rw = UIO_WRITE;
        auio.uio_resid = maxsize;
        auio.uio_td = NULL;
        error = VOP_WRITE(vp, &auio, ioflags, proc0.p_ucred);
        mycpu->gd_cnt.v_vnodeout++;
        mycpu->gd_cnt.v_vnodepgsout += ncount;

        if (error) {
                krateprintf(&vbadrate,
                            "vnode_pager_putpages: I/O error %d\n", error);
        }
        if (auio.uio_resid) {
                krateprintf(&vresrate,
                            "vnode_pager_putpages: residual I/O %zd at %lu\n",
                            auio.uio_resid, (u_long)m[0]->pindex);
        }
        if (error == 0) {
                for (i = 0; i < ncount; i++) {
                        rtvals[i] = VM_PAGER_OK;
                        vm_page_undirty(m[i]);
                }
        }
        return rtvals[0];
}

struct vnode *
vnode_pager_lock(vm_object_t object)
{
        struct thread *td = curthread;  /* XXX */
        int error;

        for (; object != NULL; object = object->backing_object) {
                if (object->type != OBJT_VNODE)
                        continue;
                if (object->flags & OBJ_DEAD)
                        return NULL;

                for (;;) {
                        struct vnode *vp = object->handle;
                        error = vget(vp, LK_SHARED | LK_RETRY | LK_CANRECURSE);
                        if (error == 0) {
                                if (object->handle != vp) {
                                        vput(vp);
                                        continue;
                                }
                                return (vp);
                        }
                        if ((object->flags & OBJ_DEAD) ||
                            (object->type != OBJT_VNODE)) {
                                return NULL;
                        }
                        kprintf("vnode_pager_lock: vp %p error %d lockstatus %d, retrying\n", vp, error, lockstatus(&vp->v_lock, td));
                        tsleep(object->handle, 0, "vnpgrl", hz);
                }
        }
        return NULL;
}