/*      $NetBSD: tmpfs_vfsops.c,v 1.78 2022/11/10 10:54:14 hannken Exp $        */

/*
* Copyright (c) 2005, 2006, 2007 The NetBSD Foundation, Inc.
* All rights reserved.
*
* This code is derived from software contributed to The NetBSD Foundation
* by Julio M. Merino Vidal, developed as part of Google's Summer of Code
* 2005 program.
*
* 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.
*
* THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. 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 FOUNDATION 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.
*/

/*
* Efficient memory file system.
*
* tmpfs is a file system that uses NetBSD's virtual memory sub-system
* (the well-known UVM) to store file data and metadata in an efficient
* way.  This means that it does not follow the structure of an on-disk
* file system because it simply does not need to.  Instead, it uses
* memory-specific data structures and algorithms to automatically
* allocate and release resources.
*/

#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: tmpfs_vfsops.c,v 1.78 2022/11/10 10:54:14 hannken Exp $");

#include <sys/param.h>
#include <sys/atomic.h>
#include <sys/types.h>
#include <sys/kmem.h>
#include <sys/mount.h>
#include <sys/stat.h>
#include <sys/systm.h>
#include <sys/vnode.h>
#include <sys/kauth.h>
#include <sys/module.h>

#include <miscfs/genfs/genfs.h>
#include <fs/tmpfs/tmpfs.h>
#include <fs/tmpfs/tmpfs_args.h>

MODULE(MODULE_CLASS_VFS, tmpfs, NULL);

struct pool     tmpfs_dirent_pool;
struct pool     tmpfs_node_pool;

void
tmpfs_init(void)
{

       pool_init(&tmpfs_dirent_pool, sizeof(tmpfs_dirent_t), 0, 0, 0,
           "tmpfs_dirent", &pool_allocator_nointr, IPL_NONE);
       pool_init(&tmpfs_node_pool, sizeof(tmpfs_node_t), 0, 0, 0,
           "tmpfs_node", &pool_allocator_nointr, IPL_NONE);
}

void
tmpfs_done(void)
{

       pool_destroy(&tmpfs_dirent_pool);
       pool_destroy(&tmpfs_node_pool);
}

int
tmpfs_mount(struct mount *mp, const char *path, void *data, size_t *data_len)
{
       struct tmpfs_args *args = data;
       tmpfs_mount_t *tmp;
       tmpfs_node_t *root;
       struct vattr va;
       struct vnode *vp;
       uint64_t memlimit;
       ino_t nodes;
       int error, flags;
       bool set_memlimit;
       bool set_nodes;

       if (args == NULL)
               return EINVAL;

       /* Validate the version. */
       if (*data_len < sizeof(*args) ||
           args->ta_version != TMPFS_ARGS_VERSION)
               return EINVAL;

       /* Handle retrieval of mount point arguments. */
       if (mp->mnt_flag & MNT_GETARGS) {
               if (mp->mnt_data == NULL)
                       return EIO;
               tmp = VFS_TO_TMPFS(mp);

               args->ta_version = TMPFS_ARGS_VERSION;
               args->ta_nodes_max = tmp->tm_nodes_max;
               args->ta_size_max = tmp->tm_mem_limit;

               root = tmp->tm_root;
               args->ta_root_uid = root->tn_uid;
               args->ta_root_gid = root->tn_gid;
               args->ta_root_mode = root->tn_mode;

               *data_len = sizeof(*args);
               return 0;
       }


       /* Prohibit mounts if there is not enough memory. */
       if (tmpfs_mem_info(true) < uvmexp.freetarg)
               return EINVAL;

       /* Check for invalid uid and gid arguments */
       if (args->ta_root_uid == VNOVAL || args->ta_root_gid == VNOVAL)
               return EINVAL;

       /* Get the memory usage limit for this file-system. */
       if (args->ta_size_max < PAGE_SIZE) {
               memlimit = UINT64_MAX;
               set_memlimit = false;
       } else {
               memlimit = args->ta_size_max;
               set_memlimit = true;
       }
       KASSERT(memlimit > 0);

       if (args->ta_nodes_max <= 3) {
               nodes = 3 + (memlimit / 1024);
               set_nodes = false;
       } else {
               nodes = args->ta_nodes_max;
               set_nodes = true;
       }
       nodes = MIN(nodes, INT_MAX);
       KASSERT(nodes >= 3);

       if (mp->mnt_flag & MNT_UPDATE) {
               tmp = VFS_TO_TMPFS(mp);
               if (set_nodes && nodes < tmp->tm_nodes_cnt)
                       return EBUSY;
               if ((mp->mnt_iflag & IMNT_WANTRDONLY)) {
                       /* Changing from read/write to read-only. */
                       flags = WRITECLOSE;
                       if ((mp->mnt_flag & MNT_FORCE))
                               flags |= FORCECLOSE;
                       error = vflush(mp, NULL, flags);
                       if (error)
                               return error;
               }
               if (set_memlimit) {
                       if ((error = tmpfs_mntmem_set(tmp, memlimit)) != 0)
                               return error;
               }
               if (set_nodes)
                       tmp->tm_nodes_max = nodes;
               root = tmp->tm_root;
               root->tn_uid = args->ta_root_uid;
               root->tn_gid = args->ta_root_gid;
               root->tn_mode = args->ta_root_mode;
               return 0;
       }

       mp->mnt_flag |= MNT_LOCAL;
       mp->mnt_stat.f_namemax = TMPFS_MAXNAMLEN;
       mp->mnt_fs_bshift = PAGE_SHIFT;
       mp->mnt_dev_bshift = DEV_BSHIFT;
       mp->mnt_iflag |= IMNT_MPSAFE | IMNT_CAN_RWTORO | IMNT_SHRLOOKUP |
           IMNT_NCLOOKUP;
       vfs_getnewfsid(mp);

       /* Allocate the tmpfs mount structure and fill it. */
       tmp = kmem_zalloc(sizeof(tmpfs_mount_t), KM_SLEEP);
       tmp->tm_nodes_max = nodes;
       tmp->tm_nodes_cnt = 0;
       LIST_INIT(&tmp->tm_nodes);

       mutex_init(&tmp->tm_lock, MUTEX_DEFAULT, IPL_NONE);
       tmpfs_mntmem_init(tmp, memlimit);
       mp->mnt_data = tmp;

       error = set_statvfs_info(path, UIO_USERSPACE, "tmpfs", UIO_SYSSPACE,
           mp->mnt_op->vfs_name, mp, curlwp);
       if (error)
               goto errout;

       /* Allocate the root node. */
       vattr_null(&va);
       va.va_type = VDIR;
       va.va_mode = args->ta_root_mode & ALLPERMS;
       va.va_uid = args->ta_root_uid;
       va.va_gid = args->ta_root_gid;
       error = vcache_new(mp, NULL, &va, NOCRED, NULL, &vp);
       if (error)
               goto errout;
       KASSERT(vp != NULL);
       root = VP_TO_TMPFS_NODE(vp);
       KASSERT(root != NULL);

       /*
        * Parent of the root inode is itself.  Also, root inode has no
        * directory entry (i.e. is never attached), thus hold an extra
        * reference (link) for it.
        */
       root->tn_links++;
       root->tn_spec.tn_dir.tn_parent = root;
       tmp->tm_root = root;
       vrele(vp);

       return 0;

errout:
       mp->mnt_data = NULL;
       tmpfs_mntmem_destroy(tmp);
       mutex_destroy(&tmp->tm_lock);
       kmem_free(tmp, sizeof(*tmp));

       return error;
}

int
tmpfs_start(struct mount *mp, int flags)
{

       return 0;
}

int
tmpfs_unmount(struct mount *mp, int mntflags)
{
       tmpfs_mount_t *tmp = VFS_TO_TMPFS(mp);
       tmpfs_node_t *node, *cnode;
       int error, flags = 0;

       /* Handle forced unmounts. */
       if (mntflags & MNT_FORCE)
               flags |= FORCECLOSE;

       /* Finalize all pending I/O. */
       error = vflush(mp, NULL, flags);
       if (error != 0)
               return error;

       /*
        * First round, detach and destroy all directory entries.
        * Also, clear the pointers to the vnodes - they are gone.
        */
       LIST_FOREACH(node, &tmp->tm_nodes, tn_entries) {
               tmpfs_dirent_t *de;

               node->tn_vnode = NULL;
               if (node->tn_type != VDIR) {
                       continue;
               }
               while ((de = TAILQ_FIRST(&node->tn_spec.tn_dir.tn_dir)) != NULL) {
                       cnode = de->td_node;
                       if (cnode && cnode != TMPFS_NODE_WHITEOUT) {
                               cnode->tn_vnode = NULL;
                       }
                       tmpfs_dir_detach(node, de);
                       tmpfs_free_dirent(tmp, de);
               }
               /* Extra virtual entry (itself for the root). */
               node->tn_links--;
       }

       /* Release the reference on root (diagnostic). */
       node = tmp->tm_root;
       node->tn_links--;

       /* Second round, destroy all inodes. */
       while ((node = LIST_FIRST(&tmp->tm_nodes)) != NULL) {
               tmpfs_free_node(tmp, node);
       }

       /* Throw away the tmpfs_mount structure. */
       tmpfs_mntmem_destroy(tmp);
       mutex_destroy(&tmp->tm_lock);
       kmem_free(tmp, sizeof(*tmp));
       mp->mnt_data = NULL;

       return 0;
}

int
tmpfs_root(struct mount *mp, int lktype, vnode_t **vpp)
{
       tmpfs_node_t *node = VFS_TO_TMPFS(mp)->tm_root;
       int error;

       error = vcache_get(mp, &node, sizeof(node), vpp);
       if (error)
               return error;
       error = vn_lock(*vpp, lktype);
       if (error) {
               vrele(*vpp);
               *vpp = NULL;
               return error;
       }

       return 0;
}

int
tmpfs_vget(struct mount *mp, ino_t ino, int lktype, vnode_t **vpp)
{

       return EOPNOTSUPP;
}

int
tmpfs_fhtovp(struct mount *mp, struct fid *fhp, int lktype, vnode_t **vpp)
{
       tmpfs_mount_t *tmp = VFS_TO_TMPFS(mp);
       tmpfs_node_t *node;
       tmpfs_fid_t tfh;
       int error;

       if (fhp->fid_len != sizeof(tmpfs_fid_t)) {
               return EINVAL;
       }
       memcpy(&tfh, fhp, sizeof(tmpfs_fid_t));

       mutex_enter(&tmp->tm_lock);
       /* XXX big oof .. use a better data structure */
       LIST_FOREACH(node, &tmp->tm_nodes, tn_entries) {
               if (node->tn_id == tfh.tf_id) {
                       /* Prevent this node from disappearing. */
                       atomic_inc_32(&node->tn_holdcount);
                       break;
               }
       }
       mutex_exit(&tmp->tm_lock);
       if (node == NULL)
               return ESTALE;

       error = vcache_get(mp, &node, sizeof(node), vpp);
       /* If this node has been reclaimed free it now. */
       if (atomic_dec_32_nv(&node->tn_holdcount) == TMPFS_NODE_RECLAIMED) {
               KASSERT(error != 0);
               tmpfs_free_node(tmp, node);
       }
       if (error)
               return (error == ENOENT ? ESTALE : error);
       error = vn_lock(*vpp, lktype);
       if (error) {
               vrele(*vpp);
               *vpp = NULL;
               return error;
       }
       if (TMPFS_NODE_GEN(node) != tfh.tf_gen) {
               vput(*vpp);
               *vpp = NULL;
               return ESTALE;
       }

       return 0;
}

int
tmpfs_vptofh(vnode_t *vp, struct fid *fhp, size_t *fh_size)
{
       tmpfs_fid_t tfh;
       tmpfs_node_t *node;

       if (*fh_size < sizeof(tmpfs_fid_t)) {
               *fh_size = sizeof(tmpfs_fid_t);
               return E2BIG;
       }
       *fh_size = sizeof(tmpfs_fid_t);
       node = VP_TO_TMPFS_NODE(vp);

       memset(&tfh, 0, sizeof(tfh));
       tfh.tf_len = sizeof(tmpfs_fid_t);
       tfh.tf_gen = TMPFS_NODE_GEN(node);
       tfh.tf_id = node->tn_id;
       memcpy(fhp, &tfh, sizeof(tfh));

       return 0;
}

int
tmpfs_statvfs(struct mount *mp, struct statvfs *sbp)
{
       tmpfs_mount_t *tmp;
       fsfilcnt_t freenodes;
       size_t avail;

       tmp = VFS_TO_TMPFS(mp);

       sbp->f_iosize = sbp->f_frsize = sbp->f_bsize = PAGE_SIZE;

       mutex_enter(&tmp->tm_acc_lock);
       avail =  tmpfs_pages_avail(tmp);
       sbp->f_blocks = (tmpfs_bytes_max(tmp) >> PAGE_SHIFT);
       sbp->f_bavail = sbp->f_bfree = avail;
       sbp->f_bresvd = 0;

       freenodes = MIN(tmp->tm_nodes_max - tmp->tm_nodes_cnt,
           avail * PAGE_SIZE / sizeof(tmpfs_node_t));

       sbp->f_files = tmp->tm_nodes_cnt + freenodes;
       sbp->f_favail = sbp->f_ffree = freenodes;
       sbp->f_fresvd = 0;
       mutex_exit(&tmp->tm_acc_lock);

       copy_statvfs_info(sbp, mp);

       return 0;
}

int
tmpfs_sync(struct mount *mp, int waitfor, kauth_cred_t uc)
{

       return 0;
}

int
tmpfs_snapshot(struct mount *mp, vnode_t *vp, struct timespec *ctime)
{

       return EOPNOTSUPP;
}

/*
* tmpfs vfs operations.
*/

extern const struct vnodeopv_desc tmpfs_fifoop_opv_desc;
extern const struct vnodeopv_desc tmpfs_specop_opv_desc;
extern const struct vnodeopv_desc tmpfs_vnodeop_opv_desc;

const struct vnodeopv_desc * const tmpfs_vnodeopv_descs[] = {
       &tmpfs_fifoop_opv_desc,
       &tmpfs_specop_opv_desc,
       &tmpfs_vnodeop_opv_desc,
       NULL,
};

struct vfsops tmpfs_vfsops = {
       .vfs_name = MOUNT_TMPFS,
       .vfs_min_mount_data = sizeof (struct tmpfs_args),
       .vfs_mount = tmpfs_mount,
       .vfs_start = tmpfs_start,
       .vfs_unmount = tmpfs_unmount,
       .vfs_root = tmpfs_root,
       .vfs_quotactl = (void *)eopnotsupp,
       .vfs_statvfs = tmpfs_statvfs,
       .vfs_sync = tmpfs_sync,
       .vfs_vget = tmpfs_vget,
       .vfs_loadvnode = tmpfs_loadvnode,
       .vfs_newvnode = tmpfs_newvnode,
       .vfs_fhtovp = tmpfs_fhtovp,
       .vfs_vptofh = tmpfs_vptofh,
       .vfs_init = tmpfs_init,
       .vfs_done = tmpfs_done,
       .vfs_snapshot = tmpfs_snapshot,
       .vfs_extattrctl = vfs_stdextattrctl,
       .vfs_suspendctl = genfs_suspendctl,
       .vfs_renamelock_enter = genfs_renamelock_enter,
       .vfs_renamelock_exit = genfs_renamelock_exit,
       .vfs_fsync = (void *)eopnotsupp,
       .vfs_opv_descs = tmpfs_vnodeopv_descs
};

static int
tmpfs_modcmd(modcmd_t cmd, void *arg)
{

       switch (cmd) {
       case MODULE_CMD_INIT:
               return vfs_attach(&tmpfs_vfsops);
       case MODULE_CMD_FINI:
               return vfs_detach(&tmpfs_vfsops);
       default:
               return ENOTTY;
       }
}