/*
 * CDDL HEADER START
 *
 * The contents of this file are subject to the terms of the
 * Common Development and Distribution License (the "License").
 * You may not use this file except in compliance with the License.
 *
 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
 * or http://www.opensolaris.org/os/licensing.
 * See the License for the specific language governing permissions
 * and limitations under the License.
 *
 * When distributing Covered Code, include this CDDL HEADER in each
 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
 * If applicable, add the following below this CDDL HEADER, with the
 * fields enclosed by brackets "[]" replaced with your own identifying
 * information: Portions Copyright [yyyy] [name of copyright owner]
 *
 * CDDL HEADER END
 */
/*
 * Copyright 2009 Sun Microsystems, Inc.  All rights reserved.
 * Use is subject to license terms.
 * Copyright 2015 Joyent, Inc.  All rights reserved.
 * Copyright (c) 2017 by Delphix. All rights reserved.
 */

/*
 * This is the /dev (hence, the sdev_ prefix) filesystem.
 */

#include <sys/types.h>
#include <sys/param.h>
#include <sys/sysmacros.h>
#include <sys/systm.h>
#include <sys/kmem.h>
#include <sys/time.h>
#include <sys/pathname.h>
#include <sys/vfs.h>
#include <sys/vfs_opreg.h>
#include <sys/vnode.h>
#include <sys/file.h>
#include <sys/stat.h>
#include <sys/uio.h>
#include <sys/stat.h>
#include <sys/errno.h>
#include <sys/cmn_err.h>
#include <sys/cred.h>
#include <sys/statvfs.h>
#include <sys/policy.h>
#include <sys/mount.h>
#include <sys/debug.h>
#include <sys/modctl.h>
#include <sys/mkdev.h>
#include <fs/fs_subr.h>
#include <sys/fs/sdev_impl.h>
#include <sys/fs/snode.h>
#include <sys/fs/dv_node.h>
#include <sys/sunndi.h>
#include <sys/mntent.h>
#include <sys/disp.h>

/*
 * /dev vfs operations.
 */

/*
 * globals
 */
struct sdev_data *sdev_origins; /* mount info for origins under /dev */
kmutex_t sdev_lock; /* used for mount/unmount/rename synchronization */
taskq_t *sdev_taskq = NULL;

/*
 * static
 */
static major_t devmajor;        /* the fictitious major we live on */
static major_t devminor;        /* the fictitious minor of this instance */
static struct sdev_data *sdev_mntinfo = NULL;   /* linked list of instances */

/* LINTED E_STATIC_UNUSED */            /* useful for debugging */
static struct vnode *sdev_stale_attrvp; /* stale root attrvp after remount */

static int sdev_mount(struct vfs *, struct vnode *, struct mounta *,
    struct cred *);
static int sdev_unmount(struct vfs *, int, struct cred *);
static int sdev_root(struct vfs *, struct vnode **);
static int sdev_statvfs(struct vfs *, struct statvfs64 *);
static void sdev_insert_mntinfo(struct sdev_data *);
static int devinit(int, char *);

static vfsdef_t sdev_vfssw = {
        VFSDEF_VERSION,
        "dev",          /* type name string */
        devinit,        /* init routine */
        VSW_CANREMOUNT, /* flags */
        NULL            /* mount options table prototype */
};


/*
 * Module linkage information
 */
static struct modlfs modlfs = {
        &mod_fsops, "/dev filesystem", &sdev_vfssw
};

static struct modlinkage modlinkage = {
        MODREV_1, (void *)&modlfs, NULL
};

int
_init(void)
{
        int e;

        mutex_init(&sdev_lock, NULL, MUTEX_DEFAULT, NULL);
        sdev_node_cache_init();
        sdev_devfsadm_lockinit();
        if ((e = mod_install(&modlinkage)) != 0) {
                sdev_devfsadm_lockdestroy();
                sdev_node_cache_fini();
                mutex_destroy(&sdev_lock);
                return (e);
        }
        return (0);
}

/*
 * dev module remained loaded for the global /dev instance
 */
int
_fini(void)
{
        return (EBUSY);
}

int
_info(struct modinfo *modinfop)
{
        return (mod_info(&modlinkage, modinfop));
}

/*ARGSUSED*/
static int
devinit(int fstype, char *name)
{
        static const fs_operation_def_t dev_vfsops_tbl[] = {
                VFSNAME_MOUNT,          { .vfs_mount = sdev_mount },
                VFSNAME_UNMOUNT,        { .vfs_unmount = sdev_unmount },
                VFSNAME_ROOT,           { .vfs_root = sdev_root },
                VFSNAME_STATVFS,        { .vfs_statvfs = sdev_statvfs },
                NULL,                   NULL
        };

        int     error;
        extern major_t getudev(void);

        devtype = fstype;

        error = vfs_setfsops(fstype, dev_vfsops_tbl, NULL);
        if (error != 0) {
                cmn_err(CE_WARN, "devinit: bad vfs ops tbl");
                return (error);
        }

        error = vn_make_ops("dev", sdev_vnodeops_tbl, &sdev_vnodeops);
        if (error != 0) {
                (void) vfs_freevfsops_by_type(fstype);
                cmn_err(CE_WARN, "devinit: bad vnode ops tbl");
                return (error);
        }

        if ((devmajor = getudev()) == (major_t)-1) {
                cmn_err(CE_WARN, "%s: can't get unique dev", sdev_vfssw.name);
                return (ENXIO);
        }

        if (sdev_plugin_init() != 0) {
                cmn_err(CE_WARN, "%s: failed to set init plugin subsystem",
                    sdev_vfssw.name);
                return (EIO);
        }

        /* initialize negative cache */
        sdev_ncache_init();

        return (0);
}

/*
 * Both mount point and backing store directory name are
 * passed in from userland
 */
static int
sdev_mount(struct vfs *vfsp, struct vnode *mvp, struct mounta *uap,
    struct cred *cr)
{
        struct sdev_data *sdev_data;
        struct vnode *avp;
        struct sdev_node *dv;
        struct sdev_mountargs *args = NULL;
        int     error = 0;
        dev_t   devdev;

        /*
         * security check
         */
        if ((secpolicy_fs_mount(cr, mvp, vfsp) != 0) ||
            (secpolicy_sys_devices(cr) != 0))
                return (EPERM);

        /*
         * Sanity check the mount point
         */
        if (mvp->v_type != VDIR)
                return (ENOTDIR);

        /*
         * Sanity Check for overlay mount.
         */
        mutex_enter(&mvp->v_lock);
        if ((uap->flags & MS_OVERLAY) == 0 &&
            (uap->flags & MS_REMOUNT) == 0 &&
            (mvp->v_count > 1 || (mvp->v_flag & VROOT))) {
                mutex_exit(&mvp->v_lock);
                return (EBUSY);
        }
        mutex_exit(&mvp->v_lock);

        args = kmem_zalloc(sizeof (*args), KM_SLEEP);

        if ((uap->flags & MS_DATA) &&
            (uap->datalen != 0 && uap->dataptr != NULL)) {
                /* copy in the arguments */
                if (error = sdev_copyin_mountargs(uap, args))
                        goto cleanup;
        }

        /*
         * Sanity check the backing store
         */
        if (args->sdev_attrdir) {
                /* user supplied an attribute store */
                if (error = lookupname((char *)(uintptr_t)args->sdev_attrdir,
                    UIO_USERSPACE, FOLLOW, NULLVPP, &avp)) {
                        cmn_err(CE_NOTE, "/dev fs: lookup on attribute "
                            "directory %s failed",
                            (char *)(uintptr_t)args->sdev_attrdir);
                        goto cleanup;
                }

                if (avp->v_type != VDIR) {
                        VN_RELE(avp);
                        error = ENOTDIR;
                        goto cleanup;
                }
        } else {
                /* use mountp as the attribute store */
                avp = mvp;
                VN_HOLD(avp);
        }

        mutex_enter(&sdev_lock);

        /*
         * Check that the taskq has been created. We can't do this in our
         * _init or devinit because they run too early for ddi_taskq_create.
         */
        if (sdev_taskq == NULL) {
                sdev_taskq = taskq_create("sdev", 1, minclsyspri, 1, 1, 0);
                if (sdev_taskq == NULL) {
                        error = ENOMEM;
                        mutex_exit(&sdev_lock);
                        VN_RELE(avp);
                        goto cleanup;
                }
        }

        /*
         * handling installation
         */
        if (uap->flags & MS_REMOUNT) {
                sdev_data = (struct sdev_data *)vfsp->vfs_data;
                ASSERT(sdev_data);

                dv = sdev_data->sdev_root;
                ASSERT(dv == dv->sdev_dotdot);

                /*
                 * mark all existing sdev_nodes (except root node) stale
                 */
                sdev_stale(dv);

                /* Reset previous mountargs */
                if (sdev_data->sdev_mountargs) {
                        kmem_free(sdev_data->sdev_mountargs,
                            sizeof (struct sdev_mountargs));
                }
                sdev_data->sdev_mountargs = args;
                args = NULL;            /* so it won't be freed below */

                sdev_stale_attrvp = dv->sdev_attrvp;
                dv->sdev_attrvp = avp;
                vfsp->vfs_mtime = ddi_get_time();

                mutex_exit(&sdev_lock);
                goto cleanup;                           /* we're done */
        }

        /*
         * Create and initialize the vfs-private data.
         */
        devdev = makedevice(devmajor, devminor);
        while (vfs_devismounted(devdev)) {
                devminor = (devminor + 1) & MAXMIN32;

                /*
                 * All the minor numbers are used up.
                 */
                if (devminor == 0) {
                        mutex_exit(&sdev_lock);
                        VN_RELE(avp);
                        error = ENODEV;
                        goto cleanup;
                }

                devdev = makedevice(devmajor, devminor);
        }

        dv = sdev_mkroot(vfsp, devdev, mvp, avp, cr);
        sdev_data = kmem_zalloc(sizeof (struct sdev_data), KM_SLEEP);
        vfsp->vfs_dev = devdev;
        vfsp->vfs_data = (caddr_t)sdev_data;
        vfsp->vfs_fstype = devtype;
        vfsp->vfs_bsize = DEV_BSIZE;
        vfsp->vfs_mtime = ddi_get_time();
        vfs_make_fsid(&vfsp->vfs_fsid, vfsp->vfs_dev, devtype);

        ASSERT(dv == dv->sdev_dotdot);

        sdev_data->sdev_vfsp = vfsp;
        sdev_data->sdev_root = dv;
        sdev_data->sdev_mountargs = args;

        /* get acl flavor from attribute dir */
        if (VOP_PATHCONF(avp, _PC_ACL_ENABLED, &sdev_data->sdev_acl_flavor,
            kcred, NULL) != 0 || sdev_data->sdev_acl_flavor == 0)
                sdev_data->sdev_acl_flavor = _ACL_ACLENT_ENABLED;

        args = NULL;                    /* so it won't be freed below */
        sdev_insert_mntinfo(sdev_data);
        mutex_exit(&sdev_lock);

        if (!SDEV_IS_GLOBAL(dv)) {
                ASSERT(sdev_origins);
                dv->sdev_flags &= ~SDEV_GLOBAL;
                dv->sdev_origin = sdev_origins->sdev_root;
                SDEV_HOLD(dv->sdev_origin);
        } else {
                sdev_ncache_setup();
                rw_enter(&dv->sdev_contents, RW_WRITER);
                sdev_filldir_dynamic(dv);
                rw_exit(&dv->sdev_contents);
        }

        sdev_update_timestamps(dv->sdev_attrvp,
            cr, AT_CTIME|AT_MTIME|AT_ATIME);

cleanup:
        if (args)
                kmem_free(args, sizeof (*args));
        return (error);
}

/*
 * unmounting the non-global /dev instances, e.g. when deleting a Kevlar zone.
 */
static int
sdev_unmount(struct vfs *vfsp, int flag, struct cred *cr)
{
        struct sdev_node *dv;
        int error;
        struct sdev_data *sdev_data, *prev, *next;

        /*
         * enforce the security policies
         */
        if ((secpolicy_fs_unmount(cr, vfsp) != 0) ||
            (secpolicy_sys_devices(cr) != 0))
                return (EPERM);

        if (flag & MS_FORCE)
                return (ENOTSUP);

        mutex_enter(&sdev_lock);
        dv = VFSTOSDEVFS(vfsp)->sdev_root;
        ASSERT(dv == dv->sdev_dotdot);
        if (SDEVTOV(dv)->v_count > 1) {
                mutex_exit(&sdev_lock);
                return (EBUSY);
        }

        /*
         * global instance remains mounted
         */
        if (SDEV_IS_GLOBAL(dv)) {
                mutex_exit(&sdev_lock);
                return (EBUSY);
        }
        mutex_exit(&sdev_lock);

        /* verify the v_count */
        if ((error = sdev_cleandir(dv, NULL, 0)) != 0) {
                return (error);
        }
        ASSERT(SDEVTOV(dv)->v_count == 1);

        /* release hold on root node and destroy it */
        SDEV_RELE(dv);
        dv->sdev_nlink -= 2;
        sdev_nodedestroy(dv, 0);

        sdev_data = (struct sdev_data *)vfsp->vfs_data;
        vfsp->vfs_data = (caddr_t)0;

        /*
         * XXX separate it into sdev_delete_mntinfo() if useful
         */
        mutex_enter(&sdev_lock);
        prev = sdev_data->sdev_prev;
        next = sdev_data->sdev_next;
        if (prev)
                prev->sdev_next = next;
        else
                sdev_mntinfo = next;
        if (next)
                next->sdev_prev = prev;
        mutex_exit(&sdev_lock);

        if (sdev_data->sdev_mountargs) {
                kmem_free(sdev_data->sdev_mountargs,
                    sizeof (struct sdev_mountargs));
        }
        kmem_free(sdev_data, sizeof (struct sdev_data));
        return (0);
}

/*
 * return root vnode for given vfs
 */
static int
sdev_root(struct vfs *vfsp, struct vnode **vpp)
{
        *vpp = SDEVTOV(VFSTOSDEVFS(vfsp)->sdev_root);
        VN_HOLD(*vpp);
        return (0);
}

/*
 * return 'generic superblock' information to userland.
 *
 * not much that we can usefully admit to here
 */
static int
sdev_statvfs(struct vfs *vfsp, struct statvfs64 *sbp)
{
        dev32_t d32;

        bzero(sbp, sizeof (*sbp));
        sbp->f_frsize = sbp->f_bsize = vfsp->vfs_bsize;
        sbp->f_files = kmem_cache_stat(sdev_node_cache, "alloc");

        /* no illusions that free/avail files is relevant to dev */
        sbp->f_ffree = 0;
        sbp->f_favail = 0;

        /* no illusions that blocks are relevant to devfs */
        sbp->f_bfree = 0;
        sbp->f_bavail = 0;
        sbp->f_blocks = 0;

        (void) cmpldev(&d32, vfsp->vfs_dev);
        sbp->f_fsid = d32;
        (void) strcpy(sbp->f_basetype, vfssw[devtype].vsw_name);
        sbp->f_flag = vf_to_stf(vfsp->vfs_flag);
        sbp->f_namemax = MAXNAMELEN - 1;
        (void) strcpy(sbp->f_fstr, "dev");

        return (0);
}

static void
sdev_insert_mntinfo(struct sdev_data *data)
{
        ASSERT(mutex_owned(&sdev_lock));
        data->sdev_next = sdev_mntinfo;
        data->sdev_prev = NULL;
        if (sdev_mntinfo) {
                sdev_mntinfo->sdev_prev = data;
        } else {
                sdev_origins = data;
        }
        sdev_mntinfo = data;
}

struct sdev_data *
sdev_find_mntinfo(char *mntpt)
{
        struct sdev_data *mntinfo;

        mutex_enter(&sdev_lock);
        mntinfo = sdev_mntinfo;
        while (mntinfo) {
                if (strcmp(mntpt, mntinfo->sdev_root->sdev_name) == 0) {
                        VN_HOLD(SDEVTOV(mntinfo->sdev_root));
                        break;
                }
                mntinfo = mntinfo->sdev_next;
        }
        mutex_exit(&sdev_lock);
        return (mntinfo);
}

void
sdev_mntinfo_rele(struct sdev_data *mntinfo)
{
        vnode_t *vp;

        mutex_enter(&sdev_lock);
        vp = SDEVTOV(mntinfo->sdev_root);
        mutex_enter(&vp->v_lock);
        VN_RELE_LOCKED(vp);
        mutex_exit(&vp->v_lock);
        mutex_exit(&sdev_lock);
}

void
sdev_mnt_walk(void (*func)(struct sdev_node *, void *), void *arg)
{
        struct sdev_data *mntinfo;

        mutex_enter(&sdev_lock);
        mntinfo = sdev_mntinfo;
        while (mntinfo != NULL) {
                func(mntinfo->sdev_root, arg);
                mntinfo = mntinfo->sdev_next;
        }
        mutex_exit(&sdev_lock);
}