kernel - Fix excessive call stack depth on stuck interrupt

[dragonfly.git] / sys / kern / subr_diskslice.c

/*-
 * Copyright (c) 1994 Bruce D. Evans.
 * All rights reserved.
 *
 * Copyright (c) 1990 The Regents of the University of California.
 * All rights reserved.
 *
 * This code is derived from software contributed to Berkeley by
 * William Jolitz.
 *
 * Copyright (c) 1982, 1986, 1988 Regents of the University of California.
 * 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. 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. 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: @(#)wd.c  7.2 (Berkeley) 5/9/91
 *      from: wd.c,v 1.55 1994/10/22 01:57:12 phk Exp $
 *      from: @(#)ufs_disksubr.c        7.16 (Berkeley) 5/4/91
 *      from: ufs_disksubr.c,v 1.8 1994/06/07 01:21:39 phk Exp $
 * $FreeBSD: src/sys/kern/subr_diskslice.c,v 1.82.2.6 2001/07/24 09:49:41 dd Exp $
 */

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/buf.h>
#include <sys/conf.h>
#include <sys/disklabel.h>
#include <sys/disklabel32.h>
#include <sys/disklabel64.h>
#include <sys/diskslice.h>
#include <sys/disk.h>
#include <sys/diskmbr.h>
#include <sys/fcntl.h>
#include <sys/malloc.h>
#include <sys/stat.h>
#include <sys/syslog.h>
#include <sys/proc.h>
#include <sys/vnode.h>
#include <sys/device.h>
#include <sys/thread2.h>

#include <vfs/ufs/dinode.h>     /* XXX used only for fs.h */
#include <vfs/ufs/fs.h>         /* XXX used only to get BBSIZE/SBSIZE */
#include <sys/devfs.h>

static int  dsreadandsetlabel(cdev_t dev, u_int flags,
                           struct diskslices *ssp, struct diskslice *sp,
                           struct disk_info *info);
static void free_ds_label (struct diskslices *ssp, int slice);
static void set_ds_label (struct diskslices *ssp, int slice, disklabel_t lp,
                           disklabel_ops_t ops);
static void set_ds_wlabel (struct diskslices *ssp, int slice, int wlabel);

/*
 * Determine the size of the transfer, and make sure it is
 * within the boundaries of the partition. Adjust transfer
 * if needed, and signal errors or early completion.
 *
 * XXX TODO:
 *      o Split buffers that are too big for the device.
 *      o Check for overflow.
 *      o Finish cleaning this up.
 *
 * This function returns 1 on success, 0 if transfer equates
 * to EOF (end of disk) or -1 on failure.  The appropriate 
 * 'errno' value is also set in bp->b_error and bp->b_flags
 * is marked with B_ERROR.
 */
struct bio *
dscheck(cdev_t dev, struct bio *bio, struct diskslices *ssp)
{
        struct buf *bp = bio->bio_buf;
        struct bio *nbio;
        disklabel_t lp;
        disklabel_ops_t ops;
        long nsec;
        u_int64_t secno;
        u_int64_t endsecno;
        u_int64_t slicerel_secno;
        struct diskslice *sp;
        u_int32_t part;
        u_int32_t slice;
        int shift;
        int mask;

        slice = dkslice(dev);
        part  = dkpart(dev);

        if (bio->bio_offset < 0) {
                kprintf("dscheck(%s): negative bio_offset %lld\n", 
                        devtoname(dev), (long long)bio->bio_offset);
                goto bad;
        }
        if (slice >= ssp->dss_nslices) {
                kprintf("dscheck(%s): slice too large %d/%d\n",
                        devtoname(dev), slice, ssp->dss_nslices);
                goto bad;
        }
        sp = &ssp->dss_slices[slice];
        /*
         * Calculate secno and nsec
         */
        if (ssp->dss_secmult == 1) {
                shift = DEV_BSHIFT;
                goto doshift;
        } else if (ssp->dss_secshift != -1) {
                shift = DEV_BSHIFT + ssp->dss_secshift;
doshift:
                mask = (1 << shift) - 1;
                if ((int)bp->b_bcount & mask)
                        goto bad_bcount;
                if ((int)bio->bio_offset & mask)
                        goto bad_blkno;
                secno = bio->bio_offset >> shift;
                nsec = bp->b_bcount >> shift;
        } else {
                if (bp->b_bcount % ssp->dss_secsize)
                        goto bad_bcount;
                if (bio->bio_offset % ssp->dss_secsize)
                        goto bad_blkno;
                secno = bio->bio_offset / ssp->dss_secsize;
                nsec = bp->b_bcount / ssp->dss_secsize;
        }

        /*
         * Calculate slice-relative sector number end slice-relative
         * limit.
         */
        if (slice == WHOLE_DISK_SLICE) {
                /*
                 * Labels have not been allowed on whole-disks for a while.
                 * This really puts the nail in the coffin.
                 *
                 * Accesses to the WHOLE_DISK_SLICE do not use a disklabel
                 * and partition numbers are special-cased.  Currently numbers
                 * less then 128 are not allowed.  Partition numbers >= 128
                 * are encoded in the high 8 bits of the 64 bit buffer offset
                 * and are fed directly through to the device with no
                 * further interpretation.  In particular, no sector
                 * translation interpretation should occur because the
                 * sector size for the special raw access may not be the
                 * same as the nominal sector size for the device.
                 */
                lp.opaque = NULL;
                if (part < 128) {
                        kprintf("dscheck(%s): illegal partition number (%d) "
                                "for WHOLE_DISK_SLICE access\n",
                                devtoname(dev), part);
                        goto bad;
                } else if (part != WHOLE_SLICE_PART) {
                        nbio = push_bio(bio);
                        nbio->bio_offset = bio->bio_offset |
                                           (u_int64_t)part << 56;
                        return(nbio);
                } else {
                        /*
                         * If writing to the raw disk request a
                         * reprobe on the last close.
                         */
                        if (bp->b_cmd == BUF_CMD_WRITE)
                                sp->ds_flags |= DSF_REPROBE;
                }

                /*
                 * sp->ds_size is for the whole disk in the WHOLE_DISK_SLICE,
                 * there are no reserved areas.
                 */
                endsecno = sp->ds_size;
                slicerel_secno = secno;
        } else if (part == WHOLE_SLICE_PART) {
                /* 
                 * NOTE! opens on a whole-slice partition will not attempt
                 * to read a disklabel in, so there may not be an in-core
                 * disklabel even if there is one on the disk.
                 */
                endsecno = sp->ds_size;
                slicerel_secno = secno;
        } else if ((lp = sp->ds_label).opaque != NULL) {
                /*
                 * A label is present, extract the partition.  Snooping of
                 * the disklabel is not supported even if accessible.  Of
                 * course, the reserved area is still write protected.
                 */
                ops = sp->ds_ops;
                if (ops->op_getpartbounds(ssp, lp, part,
                                          &slicerel_secno, &endsecno)) {
                        kprintf("dscheck(%s): partition %d out of bounds\n",
                                devtoname(dev), part);
                        goto bad;
                }
                slicerel_secno += secno;
        } else {
                /*
                 * Attempt to access partition when no disklabel present
                 */
                kprintf("dscheck(%s): attempt to access non-existent partition\n",
                        devtoname(dev));
                goto bad;
        }

        /*
         * Disallow writes to reserved areas unless ds_wlabel allows it.
         * If the reserved area is written to request a reprobe of the
         * disklabel when the slice is closed.
         */
        if (slicerel_secno < sp->ds_reserved && nsec &&
            bp->b_cmd == BUF_CMD_WRITE) {
                if (sp->ds_wlabel == 0) {
                        bp->b_error = EROFS;
                        goto error;
                }
                sp->ds_flags |= DSF_REPROBE;
        }

        /*
         * If we get here, bio_offset must be on a block boundary and
         * the sector size must be a power of 2.
         */
        if ((bio->bio_offset & (ssp->dss_secsize - 1)) ||
            (ssp->dss_secsize ^ (ssp->dss_secsize - 1)) !=
            ((ssp->dss_secsize << 1) - 1)) {
                kprintf("%s: invalid BIO offset, not sector aligned or"
                        " invalid sector size (not power of 2) %08llx %d\n",
                        devtoname(dev), (long long)bio->bio_offset,
                        ssp->dss_secsize);
                goto bad;
        }

        /*
         * EOF handling
         */
        if (secno + nsec > endsecno) {
                /*
                 * Return an error if beyond the end of the disk, or
                 * if B_BNOCLIP is set.  Tell the system that we do not
                 * need to keep the buffer around.
                 */
                if (secno > endsecno || (bp->b_flags & B_BNOCLIP))
                        goto bad;

                /*
                 * If exactly at end of disk, return an EOF.  Throw away
                 * the buffer contents, if any, by setting B_INVAL.
                 */
                if (secno == endsecno) {
                        bp->b_resid = bp->b_bcount;
                        bp->b_flags |= B_INVAL;
                        goto done;
                }

                /*
                 * Else truncate
                 */
                nsec = endsecno - secno;
                bp->b_bcount = nsec * ssp->dss_secsize;
        }

        nbio = push_bio(bio);
        nbio->bio_offset = (off_t)(sp->ds_offset + slicerel_secno) * 
                           ssp->dss_secsize;
        return (nbio);

bad_bcount:
        kprintf(
        "dscheck(%s): b_bcount %d is not on a sector boundary (ssize %d)\n",
            devtoname(dev), bp->b_bcount, ssp->dss_secsize);
        goto bad;

bad_blkno:
        kprintf(
        "dscheck(%s): bio_offset %lld is not on a sector boundary (ssize %d)\n",
            devtoname(dev), (long long)bio->bio_offset, ssp->dss_secsize);
bad:
        bp->b_error = EINVAL;
        /* fall through */
error:
        /*
         * Terminate the I/O with a ranging error.  Since the buffer is
         * either illegal or beyond the file EOF, mark it B_INVAL as well.
         */
        bp->b_resid = bp->b_bcount;
        bp->b_flags |= B_ERROR | B_INVAL;
done:
        /*
         * Caller must biodone() the originally passed bio if NULL is
         * returned.
         */
        return (NULL);
}

/*
 * dsclose() - close a cooked disk slice.
 *
 * WARNING!  The passed diskslices and related diskslice structures may
 *           be invalidated or replaced by this function, callers must
 *           reload from the disk structure for continued access.
 */
void
dsclose(cdev_t dev, int mode, struct diskslices *ssp)
{
        u_int32_t part;
        u_int32_t slice;
        struct diskslice *sp;

        slice = dkslice(dev);
        part  = dkpart(dev);
        if (slice < ssp->dss_nslices) {
                sp = &ssp->dss_slices[slice];
                dsclrmask(sp, part);
                if (sp->ds_flags & DSF_REPROBE) {
                        sp->ds_flags &= ~DSF_REPROBE;
                        if (slice == WHOLE_DISK_SLICE) {
                                disk_msg_send_sync(DISK_DISK_REPROBE,
                                                   dev->si_disk, NULL);
                                devfs_config();
                        } else {
                                disk_msg_send_sync(DISK_SLICE_REPROBE,
                                                   dev->si_disk, sp);
                                devfs_config();
                        }
                        /* ssp and sp may both be invalid after reprobe */
                }
        }
}

void
dsgone(struct diskslices **sspp)
{
        int slice;
        struct diskslices *ssp;

        if ((ssp = *sspp) != NULL) {
                for (slice = 0; slice < ssp->dss_nslices; slice++)
                        free_ds_label(ssp, slice);
                kfree(ssp, M_DEVBUF);
                *sspp = NULL;
        }
}

/*
 * For the "write" commands (DIOCSDINFO and DIOCWDINFO), this
 * is subject to the same restriction as dsopen().
 */
int
dsioctl(cdev_t dev, u_long cmd, caddr_t data, int flags,
        struct diskslices **sspp, struct disk_info *info)
{
        int error;
        disklabel_t lp;
        disklabel_t lptmp;
        disklabel_ops_t ops;
        int old_wlabel;
        u_int32_t openmask[DKMAXPARTITIONS/(sizeof(u_int32_t)*8)];
        int part;
        int slice;
        struct diskslice *sp;
        struct diskslices *ssp;

        slice = dkslice(dev);
        part = dkpart(dev);
        ssp = *sspp;
        if (ssp == NULL)
                return (EINVAL);
        if (slice >= ssp->dss_nslices)
                return (EINVAL);
        sp = &ssp->dss_slices[slice];
        lp = sp->ds_label;
        ops = sp->ds_ops;       /* may be NULL if no label */

        switch (cmd) {
        case DIOCGDVIRGIN32:
                ops = &disklabel32_ops;
                /* fall through */
        case DIOCGDVIRGIN64:
                if (cmd != DIOCGDVIRGIN32)
                        ops = &disklabel64_ops;
                /*
                 * You can only retrieve a virgin disklabel on the whole
                 * disk slice or whole-slice partition.
                 */
                if (slice != WHOLE_DISK_SLICE &&
                    part != WHOLE_SLICE_PART) {
                        return(EINVAL);
                }

                lp.opaque = data;
                ops->op_makevirginlabel(lp, ssp, sp, info);
                return (0);

        case DIOCGDINFO32:
        case DIOCGDINFO64:
                /*
                 * You can only retrieve a disklabel on the whole
                 * slice partition.
                 *
                 * We do not support labels directly on whole-disks
                 * any more (that is, disks without slices), unless the
                 * device driver has asked for a compatible label (e.g.
                 * for a CD) to allow booting off of storage that is
                 * otherwise unlabeled.
                 */
                error = 0;
                if (part != WHOLE_SLICE_PART)
                        return(EINVAL);
                if (slice == WHOLE_DISK_SLICE &&
                    (info->d_dsflags & DSO_COMPATLABEL) == 0) {
                        return (ENODEV);
                }
                if (sp->ds_label.opaque == NULL) {
                        error = dsreadandsetlabel(dev, info->d_dsflags,
                                                  ssp, sp, info);
                        ops = sp->ds_ops;       /* may be NULL */
                }

                /*
                 * The type of label we found must match the type of
                 * label requested.
                 */
                if (error == 0 && IOCPARM_LEN(cmd) != ops->labelsize)
                        error = ENOATTR;
                if (error == 0)
                        bcopy(sp->ds_label.opaque, data, ops->labelsize);
                return (error);

        case DIOCGPART:
                {
                        struct partinfo *dpart = (void *)data;

                        /*
                         * The disk management layer may not have read the
                         * disklabel yet because simply opening a slice no
                         * longer 'probes' the disk that way.  Be sure we
                         * have tried.
                         *
                         * We ignore any error.
                         */
                        if (sp->ds_label.opaque == NULL &&
                            part == WHOLE_SLICE_PART &&
                            slice != WHOLE_DISK_SLICE) {
                                dsreadandsetlabel(dev, info->d_dsflags,
                                                  ssp, sp, info);
                                ops = sp->ds_ops;       /* may be NULL */
                        }

                        bzero(dpart, sizeof(*dpart));
                        dpart->media_offset   = (u_int64_t)sp->ds_offset *
                                                info->d_media_blksize;
                        dpart->media_size     = (u_int64_t)sp->ds_size *
                                                info->d_media_blksize;
                        dpart->media_blocks   = sp->ds_size;
                        dpart->media_blksize  = info->d_media_blksize;
                        dpart->reserved_blocks= sp->ds_reserved;
                        dpart->fstype_uuid = sp->ds_type_uuid;
                        dpart->storage_uuid = sp->ds_stor_uuid;

                        if (slice != WHOLE_DISK_SLICE &&
                            part != WHOLE_SLICE_PART) {
                                u_int64_t start;
                                u_int64_t blocks;
                                if (lp.opaque == NULL)
                                        return(EINVAL);
                                if (ops->op_getpartbounds(ssp, lp, part,
                                                          &start, &blocks)) {
                                        return(EINVAL);
                                }
                                ops->op_loadpartinfo(lp, part, dpart);
                                dpart->media_offset += start *
                                                       info->d_media_blksize;
                                dpart->media_size = blocks *
                                                    info->d_media_blksize;
                                dpart->media_blocks = blocks;

                                /*
                                 * partition starting sector (p_offset)
                                 * requires slice's reserved areas to be
                                 * adjusted.
                                 */
                                if (dpart->reserved_blocks > start)
                                        dpart->reserved_blocks -= start;
                                else
                                        dpart->reserved_blocks = 0;
                        }

                        /*
                         * Load remaining fields from the info structure
                         */
                        dpart->d_nheads =       info->d_nheads;
                        dpart->d_ncylinders =   info->d_ncylinders;
                        dpart->d_secpertrack =  info->d_secpertrack;
                        dpart->d_secpercyl =    info->d_secpercyl;
                }
                return (0);

        case DIOCGSLICEINFO:
                bcopy(ssp, data, (char *)&ssp->dss_slices[ssp->dss_nslices] -
                                 (char *)ssp);
                return (0);

        case DIOCSDINFO32:
                ops = &disklabel32_ops;
                /* fall through */
        case DIOCSDINFO64:
                if (cmd != DIOCSDINFO32)
                        ops = &disklabel64_ops;
                /*
                 * You can write a disklabel on the whole disk slice or
                 * whole-slice partition.
                 */
                if (slice != WHOLE_DISK_SLICE &&
                    part != WHOLE_SLICE_PART) {
                        return(EINVAL);
                }

                /*
                 * We no longer support writing disklabels directly to media
                 * without there being a slice.  Keep this as a separate
                 * conditional.
                 */
                if (slice == WHOLE_DISK_SLICE)
                        return (ENODEV);
                if (!(flags & FWRITE))
                        return (EBADF);

                /*
                 * If an existing label is present it must be the same
                 * type as the label being passed by the ioctl.
                 */
                if (sp->ds_label.opaque && sp->ds_ops != ops)
                        return (ENOATTR);

                /*
                 * Create a temporary copy of the existing label
                 * (if present) so setdisklabel can compare it against
                 * the new label.
                 */
                lp.opaque = kmalloc(ops->labelsize, M_DEVBUF, M_WAITOK);
                if (sp->ds_label.opaque == NULL)
                        bzero(lp.opaque, ops->labelsize);
                else
                        bcopy(sp->ds_label.opaque, lp.opaque, ops->labelsize);
                if (sp->ds_label.opaque == NULL) {
                        bzero(openmask, sizeof(openmask));
                } else {
                        bcopy(sp->ds_openmask, openmask, sizeof(openmask));
                }
                lptmp.opaque = data;
                error = ops->op_setdisklabel(lp, lptmp, ssp, sp, openmask);
                disk_msg_send_sync(DISK_SLICE_REPROBE, dev->si_disk, sp);
                devfs_config();
                if (error != 0) {
                        kfree(lp.opaque, M_DEVBUF);
                        return (error);
                }
                free_ds_label(ssp, slice);
                set_ds_label(ssp, slice, lp, ops);
                return (0);

        case DIOCSYNCSLICEINFO:
                /*
                 * This ioctl can only be done on the whole disk
                 */
                if (slice != WHOLE_DISK_SLICE || part != WHOLE_SLICE_PART)
                        return (EINVAL);

                if (*(int *)data == 0) {
                        for (slice = 0; slice < ssp->dss_nslices; slice++) {
                                struct diskslice *ds = &ssp->dss_slices[slice];

                                switch(dscountmask(ds)) {
                                case 0:
                                        break;
                                case 1:
                                        if (slice != WHOLE_DISK_SLICE)
                                                return (EBUSY);
                                        if (!dschkmask(ds, RAW_PART))
                                                return (EBUSY);
                                        break;
                                default:
                                        return (EBUSY);
                                }
                        }
                }

                disk_msg_send_sync(DISK_DISK_REPROBE, dev->si_disk, NULL);
                devfs_config();
                return 0;

        case DIOCWDINFO32:
        case DIOCWDINFO64:
                error = dsioctl(dev, ((cmd == DIOCWDINFO32) ?
                                        DIOCSDINFO32 : DIOCSDINFO64),
                                data, flags, &ssp, info);
                if (error == 0 && sp->ds_label.opaque == NULL)
                        error = EINVAL;
                if (part != WHOLE_SLICE_PART)
                        error = EINVAL;
                if (error != 0)
                        return (error);

                /*
                 * Allow the reserved area to be written, reload ops
                 * because the DIOCSDINFO op above may have installed
                 * a new label type.
                 */
                ops = sp->ds_ops;
                old_wlabel = sp->ds_wlabel;
                set_ds_wlabel(ssp, slice, TRUE);
                error = ops->op_writedisklabel(dev, ssp, sp, sp->ds_label);
                disk_msg_send_sync(DISK_SLICE_REPROBE, dev->si_disk, sp);
                devfs_config();
                set_ds_wlabel(ssp, slice, old_wlabel);
                /* XXX should invalidate in-core label if write failed. */
                return (error);

        case DIOCWLABEL:
                if (slice == WHOLE_DISK_SLICE)
                        return (ENODEV);
                if (!(flags & FWRITE))
                        return (EBADF);
                set_ds_wlabel(ssp, slice, *(int *)data != 0);
                return (0);

        default:
                return (ENOIOCTL);
        }
}

int
dsisopen(struct diskslices *ssp)
{
        int slice;

        if (ssp == NULL)
                return (0);
        for (slice = 0; slice < ssp->dss_nslices; slice++) {
                if (dscountmask(&ssp->dss_slices[slice]))
                        return (1);
        }
        return (0);
}

/*
 * Allocate a slices "struct" and initialize it to contain only an empty
 * compatibility slice (pointing to itself), a whole disk slice (covering
 * the disk as described by the label), and (nslices - BASE_SLICES) empty
 * slices beginning at BASE_SLICE.
 *
 * Note that the compatibility slice is no longer really a compatibility
 * slice.  It is slice 0 if a GPT label is present, and the dangerously
 * dedicated slice if no slice table otherwise exists.  Else it is 0-sized.
 */
struct diskslices *
dsmakeslicestruct(int nslices, struct disk_info *info)
{
        struct diskslice *sp;
        struct diskslices *ssp;

        ssp = kmalloc(offsetof(struct diskslices, dss_slices) +
                     nslices * sizeof *sp, M_DEVBUF, M_WAITOK);
        ssp->dss_first_bsd_slice = COMPATIBILITY_SLICE;
        ssp->dss_nslices = nslices;
        ssp->dss_oflags = 0;

        /*
         * Figure out if we can use shifts or whether we have to
         * use mod/multply to translate byte offsets into sector numbers.
         */
        if ((info->d_media_blksize ^ (info->d_media_blksize - 1)) ==
             (info->d_media_blksize << 1) - 1) {
                ssp->dss_secmult = info->d_media_blksize / DEV_BSIZE;
                if (ssp->dss_secmult & (ssp->dss_secmult - 1))
                        ssp->dss_secshift = -1;
                else
                        ssp->dss_secshift = ffs(ssp->dss_secmult) - 1;
        } else {
                ssp->dss_secmult = 0;
                ssp->dss_secshift = -1;
        }
        ssp->dss_secsize = info->d_media_blksize;
        sp = &ssp->dss_slices[0];
        bzero(sp, nslices * sizeof *sp);
        sp[WHOLE_DISK_SLICE].ds_size = info->d_media_blocks;
        return (ssp);
}

char *
dsname(cdev_t dev, int unit, int slice, int part, char *partname)
{
        return dev->si_name;
}

/*
 * This should only be called when the unit is inactive and the strategy
 * routine should not allow it to become active unless we call it.  Our
 * strategy routine must be special to allow activity.
 */
int
dsopen(cdev_t dev, int mode, u_int flags, 
       struct diskslices **sspp, struct disk_info *info)
{
        struct diskslice *sp;
        struct diskslices *ssp;
        int slice;
        int part;

        ssp = *sspp;
        dev->si_bsize_phys = info->d_media_blksize;
        slice = dkslice(dev);
        part = dkpart(dev);
        sp = &ssp->dss_slices[slice];
        dssetmask(sp, part);

        return 0;
}

/*
 * Attempt to read the disklabel.  If successful, store it in sp->ds_label.
 *
 * If we cannot read the disklabel and DSO_COMPATLABEL is set, we construct
 * a fake label covering the whole disk.
 */
static
int
dsreadandsetlabel(cdev_t dev, u_int flags,
                  struct diskslices *ssp, struct diskslice *sp,
                  struct disk_info *info)
{
        disklabel_t lp;
        disklabel_ops_t ops;
        const char *msg;
        const char *sname;
        char partname[2];
        int slice = dkslice(dev);

        /*
         * Probe the disklabel
         */
        lp.opaque = NULL;
        sname = dsname(dev, dkunit(dev), slice, WHOLE_SLICE_PART, partname);
        ops = &disklabel32_ops;
        msg = ops->op_readdisklabel(dev, sp, &lp, info);
        if (msg && strcmp(msg, "no disk label") == 0) {
                ops = &disklabel64_ops;
                msg = disklabel64_ops.op_readdisklabel(dev, sp, &lp, info);
        }

        /*
         * If we failed and COMPATLABEL is set, create a dummy disklabel.
         */
        if (msg != NULL && (flags & DSO_COMPATLABEL)) {
                msg = NULL;
                if (sp->ds_size >= 0x100000000ULL)
                        ops = &disklabel64_ops;
                else
                        ops = &disklabel32_ops;
                lp = ops->op_clone_label(info, sp);
        }
        if (msg != NULL) {
                if (sp->ds_type == DOSPTYP_386BSD /* XXX */)
                        log(LOG_WARNING, "%s: cannot find label (%s)\n",
                            sname, msg);
                if (lp.opaque)
                        kfree(lp.opaque, M_DEVBUF);
        } else {
                set_ds_label(ssp, slice, lp, ops);
                set_ds_wlabel(ssp, slice, FALSE);
        }
        return (msg ? EINVAL : 0);
}

int64_t
dssize(cdev_t dev, struct diskslices **sspp)
{
        disklabel_t lp;
        disklabel_ops_t ops;
        int part;
        int slice;
        struct diskslices *ssp;
        u_int64_t start;
        u_int64_t blocks;

        slice = dkslice(dev);
        part = dkpart(dev);
        ssp = *sspp;
        if (ssp == NULL || slice >= ssp->dss_nslices
            || !dschkmask(&ssp->dss_slices[slice], part)) {
                if (dev_dopen(dev, FREAD, S_IFCHR, proc0.p_ucred, NULL) != 0)
                        return (-1);
                dev_dclose(dev, FREAD, S_IFCHR, NULL);
                ssp = *sspp;
        }
        lp = ssp->dss_slices[slice].ds_label;
        if (part == WHOLE_SLICE_PART) {
                blocks = ssp->dss_slices[slice].ds_size;
        } else if (lp.opaque == NULL) {
                blocks = (u_int64_t)-1;
        } else {
                ops = ssp->dss_slices[slice].ds_ops;
                if (ops->op_getpartbounds(ssp, lp, part, &start, &blocks))
                        return (-1);
        }
        return ((int64_t)blocks);
}

static void
free_ds_label(struct diskslices *ssp, int slice)
{
        struct diskslice *sp;
        disklabel_t lp;

        sp = &ssp->dss_slices[slice];
        lp = sp->ds_label;
        if (lp.opaque != NULL) {
                kfree(lp.opaque, M_DEVBUF);
                lp.opaque = NULL;
                set_ds_label(ssp, slice, lp, NULL);
        }
}

static void
set_ds_label(struct diskslices *ssp, int slice,
             disklabel_t lp, disklabel_ops_t ops)
{
        struct diskslice *sp = &ssp->dss_slices[slice];

        sp->ds_label = lp;
        sp->ds_ops = ops;
        if (lp.opaque && slice != WHOLE_DISK_SLICE)
                ops->op_adjust_label_reserved(ssp, slice, sp);
        else
                sp->ds_reserved = 0;
}

static void
set_ds_wlabel(struct diskslices *ssp, int slice, int wlabel)
{
        ssp->dss_slices[slice].ds_wlabel = wlabel;
}