renamed and cleaned up export header files to match linux convention

[lunaix-os.git] / lunaix-os / kernel / fs / vfs.c

diff --git a/lunaix-os/kernel/fs/vfs.c b/lunaix-os/kernel/fs/vfs.c
index 7959fc3ffc3cce0cc58d79ec40ec777666480e5f..8bcb64430d9b2eac65f8414c6075f7dbf76f988b 100644 (file)
--- a/lunaix-os/kernel/fs/vfs.c
+++ b/lunaix-os/kernel/fs/vfs.c
@@ -40,14 +40,13 @@
                     we can do similar thing in Lunaix? A block device emulation
                     above the regular file when we mount it on.
  10. (device) device number (dev_t) allocation
                     we can do similar thing in Lunaix? A block device emulation
                     above the regular file when we mount it on.
  10. (device) device number (dev_t) allocation

-            [good idea] <class>:<subclass>:<uniq_id> composition
+            [good idea] <class>:<subclass>:<uniq_id> composition (CHECKED)

 */
 
 #include <klibc/string.h>
 #include <lunaix/foptions.h>
 #include <lunaix/fs.h>
 #include <lunaix/mm/cake.h>
 */
 
 #include <klibc/string.h>
 #include <lunaix/foptions.h>
 #include <lunaix/fs.h>
 #include <lunaix/mm/cake.h>

-#include <lunaix/mm/page.h>

 #include <lunaix/mm/valloc.h>
 #include <lunaix/process.h>
 #include <lunaix/spike.h>
 #include <lunaix/mm/valloc.h>
 #include <lunaix/process.h>
 #include <lunaix/spike.h>


@@ -56,7 +55,10 @@
 
 #include <lunaix/fs/twifs.h>
 
 
 #include <lunaix/fs/twifs.h>
 

-#include <sys/dirent_defs.h>
+#include <usr/lunaix/dirent.h>
+
+#define INODE_ACCESSED  0
+#define INODE_MODIFY    1

 
 static struct cake_pile* dnode_pile;
 static struct cake_pile* inode_pile;
 
 static struct cake_pile* dnode_pile;
 static struct cake_pile* inode_pile;


@@ -64,8 +66,7 @@ static struct cake_pile* file_pile;
 static struct cake_pile* superblock_pile;
 static struct cake_pile* fd_pile;
 
 static struct cake_pile* superblock_pile;
 static struct cake_pile* fd_pile;
 

-struct v_dnode* vfs_sysroot;
-static struct hbucket* dnode_cache;
+struct v_dnode* vfs_sysroot = NULL;

 
 struct lru_zone *dnode_lru, *inode_lru;
 
 
 struct lru_zone *dnode_lru, *inode_lru;
 


@@ -73,12 +74,6 @@ struct hstr vfs_ddot = HSTR("..", 2);
 struct hstr vfs_dot = HSTR(".", 1);
 struct hstr vfs_empty = HSTR("", 0);
 
 struct hstr vfs_dot = HSTR(".", 1);
 struct hstr vfs_empty = HSTR("", 0);
 

-struct v_superblock*
-vfs_sb_alloc();
-
-void
-vfs_sb_free(struct v_superblock* sb);
-

 static int
 __vfs_try_evict_dnode(struct lru_node* obj);
 
 static int
 __vfs_try_evict_dnode(struct lru_node* obj);
 


@@ -96,10 +91,8 @@ vfs_init()
     superblock_pile =
       cake_new_pile("sb_cache", sizeof(struct v_superblock), 1, 0);
 
     superblock_pile =
       cake_new_pile("sb_cache", sizeof(struct v_superblock), 1, 0);
 

-    dnode_cache = vzalloc(VFS_HASHTABLE_SIZE * sizeof(struct hbucket));
-
-    dnode_lru = lru_new_zone(__vfs_try_evict_dnode);
-    inode_lru = lru_new_zone(__vfs_try_evict_inode);
+    dnode_lru = lru_new_zone("vfs_dnode", __vfs_try_evict_dnode);
+    inode_lru = lru_new_zone("vfs_inode", __vfs_try_evict_inode);

 
     hstr_rehash(&vfs_ddot, HSTR_FULL_HASH);
     hstr_rehash(&vfs_dot, HSTR_FULL_HASH);
 
     hstr_rehash(&vfs_ddot, HSTR_FULL_HASH);
     hstr_rehash(&vfs_dot, HSTR_FULL_HASH);


@@ -107,24 +100,79 @@ vfs_init()
     // 创建一个根dnode。
     vfs_sysroot = vfs_d_alloc(NULL, &vfs_empty);
     vfs_sysroot->parent = vfs_sysroot;
     // 创建一个根dnode。
     vfs_sysroot = vfs_d_alloc(NULL, &vfs_empty);
     vfs_sysroot->parent = vfs_sysroot;

-    atomic_fetch_add(&vfs_sysroot->ref_count, 1);
+    
+    vfs_ref_dnode(vfs_sysroot);
+    lru_remove(dnode_lru, &vfs_sysroot->lru);

 }
 
 }
 

-inline struct hbucket*
-__dcache_hash(struct v_dnode* parent, u32_t* hash)
+void
+vfs_vncache_init(struct vncache* cache)
+{
+    cache->pool = vzalloc(VFS_HASHTABLE_SIZE * sizeof(struct hbucket));
+    rwlock_init(&cache->lock);
+}
+
+void
+vfs_vncache_free(struct vncache* cache)
+{
+    // clear all other reader/writer
+    rwlock_begin_write(&cache->lock);
+    vfree(cache->pool);
+    
+    // already freed, so as the lock
+}
+
+void
+vfs_vncache_add(struct vncache* cache, size_t key, struct hlist_node* node)

 {
 {

-    u32_t _hash = *hash;
-    // 确保低位更加随机
+    struct hbucket* slot;
+
+    cache_atomic_write(cache, 
+    {
+        slot = &cache->pool[key & VFS_HASH_MASK];
+        hlist_delete(node);
+        hlist_add(&slot->head, node);
+    });
+}
+
+static inline struct hbucket*
+__dcache_hash_nolock(struct v_dnode* parent, u32_t* hash)
+{
+    struct v_superblock* sb;
+    struct hbucket* d_cache;
+    u32_t _hash;
+
+    sb = parent->super_block;
+    
+    _hash = *hash;

     _hash = _hash ^ (_hash >> VFS_HASHBITS);
     _hash = _hash ^ (_hash >> VFS_HASHBITS);

-    // 与parent的指针值做加法，来减小碰撞的可能性。
-    _hash += (u32_t)parent;
+    _hash += (u32_t)__ptr(parent);
+

     *hash = _hash;
     *hash = _hash;

-    return &dnode_cache[_hash & VFS_HASH_MASK];
+    return &sb->d_cache.pool[_hash & VFS_HASH_MASK];
+}
+
+static inline int
+__sync_inode_nolock(struct v_inode* inode)
+{
+    pcache_commit_all(inode);
+
+    int errno = ENOTSUP;
+    if (inode->ops->sync) {
+        errno = inode->ops->sync(inode);
+    }
+
+    return errno;

 }
 
 struct v_dnode*
 vfs_dcache_lookup(struct v_dnode* parent, struct hstr* str)
 {
 }
 
 struct v_dnode*
 vfs_dcache_lookup(struct v_dnode* parent, struct hstr* str)
 {

+    u32_t hash;
+    struct hbucket* slot;
+    struct v_dnode *pos, *n;
+    struct vncache *dcache;
+ 

     if (!str->len || HSTR_EQ(str, &vfs_dot))
         return parent;
 
     if (!str->len || HSTR_EQ(str, &vfs_dot))
         return parent;
 


@@ -132,30 +180,58 @@ vfs_dcache_lookup(struct v_dnode* parent, struct hstr* str)
         return parent->parent;
     }
 
         return parent->parent;
     }
 

-    u32_t hash = str->hash;
-    struct hbucket* slot = __dcache_hash(parent, &hash);
+    hash = str->hash;
+    dcache = dnode_cache(parent);
+    
+    vncache_lock_read(dcache);

 
 

-    struct v_dnode *pos, *n;
+    slot = __dcache_hash_nolock(parent, &hash);

     hashtable_bucket_foreach(slot, pos, n, hash_list)
     {
     hashtable_bucket_foreach(slot, pos, n, hash_list)
     {

-        if (pos->name.hash == hash) {
-            return pos;
+        if (pos->name.hash != hash || pos->parent != parent) {
+            continue;

         }
         }

+
+        vncache_unlock_read(dcache);
+        return pos;

     }
     }

+
+    vncache_unlock_read(dcache);

     return NULL;
 }
 
     return NULL;
 }
 

+static void
+__vfs_touch_inode(struct v_inode* inode, const int type)
+{
+    if (type == INODE_MODIFY) {
+        inode->mtime = clock_unixtime();
+    }
+
+    else if (type == INODE_ACCESSED) {
+        inode->atime = clock_unixtime();
+    }
+
+    lru_use_one(inode_lru, &inode->lru);
+}
+

 void
 vfs_dcache_add(struct v_dnode* parent, struct v_dnode* dnode)
 {
 void
 vfs_dcache_add(struct v_dnode* parent, struct v_dnode* dnode)
 {

+    struct hbucket* bucket;
+    struct vncache* cache;
+

     assert(parent);
     assert(parent);

+    assert(locked_node(parent));

 
 

-    atomic_fetch_add(&dnode->ref_count, 1);
+    dnode->ref_count = 1;

     dnode->parent = parent;
     llist_append(&parent->children, &dnode->siblings);
 
     dnode->parent = parent;
     llist_append(&parent->children, &dnode->siblings);
 

-    struct hbucket* bucket = __dcache_hash(parent, &dnode->name.hash);
-    hlist_add(&bucket->head, &dnode->hash_list);
+    cache_atomic_write(dnode_cache(parent), 
+    {
+        bucket = __dcache_hash_nolock(parent, &dnode->name.hash);
+        hlist_add(&bucket->head, &dnode->hash_list);
+    });

 }
 
 void
 }
 
 void


@@ -166,31 +242,40 @@ vfs_dcache_remove(struct v_dnode* dnode)
 
     llist_delete(&dnode->siblings);
     llist_delete(&dnode->aka_list);
 
     llist_delete(&dnode->siblings);
     llist_delete(&dnode->aka_list);

-    hlist_delete(&dnode->hash_list);
+    lru_remove(dnode_lru, &dnode->lru);
+
+    cache_atomic_write(dnode_cache(dnode),
+    {
+        hlist_delete(&dnode->hash_list);
+    });

 
     dnode->parent = NULL;
 
     dnode->parent = NULL;

-    atomic_fetch_sub(&dnode->ref_count, 1);
+    dnode->ref_count = 0;

 }
 
 void
 vfs_dcache_rehash(struct v_dnode* new_parent, struct v_dnode* dnode)
 {
     assert(new_parent);
 }
 
 void
 vfs_dcache_rehash(struct v_dnode* new_parent, struct v_dnode* dnode)
 {
     assert(new_parent);

+    assert(locked_node(new_parent));

 
 

-    hstr_rehash(&dnode->name, HSTR_FULL_HASH);
-    vfs_dcache_remove(dnode);
-    vfs_dcache_add(new_parent, dnode);
+    dnode_atomic(dnode, 
+    {
+        hstr_rehash(&dnode->name, HSTR_FULL_HASH);
+        vfs_dcache_remove(dnode);
+        vfs_dcache_add(new_parent, dnode);
+    });

 }
 
 int
 vfs_open(struct v_dnode* dnode, struct v_file** file)
 {
 }
 
 int
 vfs_open(struct v_dnode* dnode, struct v_file** file)
 {

-    if (!dnode->inode || !dnode->inode->ops->open) {
+    struct v_inode* inode = dnode->inode;
+    
+    if (!inode || !inode->ops->open) {

         return ENOTSUP;
     }
 
         return ENOTSUP;
     }
 

-    struct v_inode* inode = dnode->inode;
-

     lock_inode(inode);
 
     struct v_file* vfile = cake_grab(file_pile);
     lock_inode(inode);
 
     struct v_file* vfile = cake_grab(file_pile);


@@ -198,10 +283,10 @@ vfs_open(struct v_dnode* dnode, struct v_file** file)
 
     vfile->dnode = dnode;
     vfile->inode = inode;
 
     vfile->dnode = dnode;
     vfile->inode = inode;

-    vfile->ref_count = ATOMIC_VAR_INIT(1);
+    vfile->ref_count = 1;

     vfile->ops = inode->default_fops;
 
     vfile->ops = inode->default_fops;
 

-    if ((inode->itype & VFS_IFFILE) && !inode->pg_cache) {
+    if (check_regfile_node(inode) && !inode->pg_cache) {

         struct pcache* pcache = vzalloc(sizeof(struct pcache));
         pcache_init(pcache);
         pcache->master = inode;
         struct pcache* pcache = vzalloc(sizeof(struct pcache));
         pcache_init(pcache);
         pcache->master = inode;


@@ -212,9 +297,8 @@ vfs_open(struct v_dnode* dnode, struct v_file** file)
     if (errno) {
         cake_release(file_pile, vfile);
     } else {
     if (errno) {
         cake_release(file_pile, vfile);
     } else {

-        atomic_fetch_add(&dnode->ref_count, 1);
+        vfs_ref_dnode(dnode);

         inode->open_count++;
         inode->open_count++;

-        mnt_mkbusy(dnode->mnt);

 
         *file = vfile;
     }
 
         *file = vfile;
     }


@@ -227,33 +311,43 @@ vfs_open(struct v_dnode* dnode, struct v_file** file)
 void
 vfs_assign_inode(struct v_dnode* assign_to, struct v_inode* inode)
 {
 void
 vfs_assign_inode(struct v_dnode* assign_to, struct v_inode* inode)
 {

+    lock_dnode(assign_to);
+

     if (assign_to->inode) {
         llist_delete(&assign_to->aka_list);
         assign_to->inode->link_count--;
     }
     if (assign_to->inode) {
         llist_delete(&assign_to->aka_list);
         assign_to->inode->link_count--;
     }

+

     llist_append(&inode->aka_dnodes, &assign_to->aka_list);
     assign_to->inode = inode;
     inode->link_count++;
     llist_append(&inode->aka_dnodes, &assign_to->aka_list);
     assign_to->inode = inode;
     inode->link_count++;

+
+    unlock_dnode(assign_to);

 }
 
 int
 vfs_link(struct v_dnode* to_link, struct v_dnode* name)
 {
     int errno;
 }
 
 int
 vfs_link(struct v_dnode* to_link, struct v_dnode* name)
 {
     int errno;

+    struct v_inode* inode;
+
+    inode = to_link->inode;

 
     if ((errno = vfs_check_writable(to_link))) {
         return errno;
     }
 
 
     if ((errno = vfs_check_writable(to_link))) {
         return errno;
     }
 

-    lock_inode(to_link->inode);
+    lock_inode(inode);
+

     if (to_link->super_block->root != name->super_block->root) {
         errno = EXDEV;
     if (to_link->super_block->root != name->super_block->root) {
         errno = EXDEV;

-    } else if (!to_link->inode->ops->link) {
+    } else if (!inode->ops->link) {

         errno = ENOTSUP;
         errno = ENOTSUP;

-    } else if (!(errno = to_link->inode->ops->link(to_link->inode, name))) {
-        vfs_assign_inode(name, to_link->inode);
+    } else if (!(errno = inode->ops->link(inode, name))) {
+        vfs_assign_inode(name, inode);

     }
     }

-    unlock_inode(to_link->inode);
+
+    unlock_inode(inode);

 
     return errno;
 }
 
     return errno;
 }


@@ -261,38 +355,55 @@ vfs_link(struct v_dnode* to_link, struct v_dnode* name)
 int
 vfs_pclose(struct v_file* file, pid_t pid)
 {
 int
 vfs_pclose(struct v_file* file, pid_t pid)
 {

+    struct v_inode* inode;

     int errno = 0;
     int errno = 0;

-    if (file->ref_count > 1) {
-        atomic_fetch_sub(&file->ref_count, 1);
-    } else if (!(errno = file->ops->close(file))) {
-        atomic_fetch_sub(&file->dnode->ref_count, 1);
-        file->inode->open_count--;
-
-        /*
-         * Prevent dead lock.
-         * This happened when process is terminated while blocking on read.
-         * In that case, the process is still holding the inode lock and it
-             will never get released.
-         * The unlocking should also include ownership check.
-         *
-         * To see why, consider two process both open the same file both with
-         * fd=x.
-         *      Process A: busy on reading x
-         *      Process B: do nothing with x
-         * Assuming that, after a very short time, process B get terminated
-         * while process A is still busy in it's reading business. By this
-         * design, the inode lock of this file x is get released by B rather
-         * than A. And this will cause a probable race condition on A if other
-         * process is writing to this file later after B exit.
-         */
-        if (mutex_on_hold(&file->inode->lock)) {
-            mutex_unlock_for(&file->inode->lock, pid);
-        }
-        mnt_chillax(file->dnode->mnt);

 
 

-        pcache_commit_all(file->inode);
-        cake_release(file_pile, file);
+    inode = file->inode;
+
+    if (vfs_check_duped_file(file)) {
+        vfs_unref_file(file);
+        return 0;
+    }
+
+    /*
+     * Prevent dead lock.
+     * This happened when process is terminated while blocking on read.
+     * In that case, the process is still holding the inode lock and it
+         will never get released.
+     * The unlocking should also include ownership check.
+     *
+     * To see why, consider two process both open the same file both with
+     * fd=x.
+     *      Process A: busy on reading x
+     *      Process B: do nothing with x
+     * Assuming that, after a very short time, process B get terminated
+     * while process A is still busy in it's reading business. By this
+     * design, the inode lock of this file x is get released by B rather
+     * than A. And this will cause a probable race condition on A if other
+     * process is writing to this file later after B exit.
+    */
+    mutex_unlock_for(&inode->lock, pid);
+
+    // now regain lock for inode syncing
+
+    lock_inode(inode);
+
+    if ((errno = file->ops->close(file))) {
+        goto done;
+    }
+
+    vfs_unref_dnode(file->dnode);
+    cake_release(file_pile, file);
+
+    pcache_commit_all(inode);
+    inode->open_count--;
+
+    if (!inode->open_count) {
+        __sync_inode_nolock(inode);

     }
     }

+
+done:
+    unlock_inode(inode);

     return errno;
 }
 
     return errno;
 }
 


@@ -308,6 +419,18 @@ vfs_free_fd(struct v_fd* fd)
     cake_release(fd_pile, fd);
 }
 
     cake_release(fd_pile, fd);
 }
 

+int
+vfs_isync(struct v_inode* inode)
+{
+    lock_inode(inode);
+
+    int errno = __sync_inode_nolock(inode);
+
+    unlock_inode(inode);
+
+    return errno;
+}
+

 int
 vfs_fsync(struct v_file* file)
 {
 int
 vfs_fsync(struct v_file* file)
 {


@@ -316,29 +439,28 @@ vfs_fsync(struct v_file* file)
         return errno;
     }
 
         return errno;
     }
 

-    lock_inode(file->inode);
-
-    pcache_commit_all(file->inode);
-
-    errno = ENOTSUP;
-    if (file->ops->sync) {
-        errno = file->ops->sync(file);
-    }
-
-    unlock_inode(file->inode);
-
-    return errno;
+    return vfs_isync(file->inode);

 }
 
 int
 vfs_alloc_fdslot(int* fd)
 {
 }
 
 int
 vfs_alloc_fdslot(int* fd)
 {

+    struct v_fdtable* fdtab;
+
+    fdtab = __current->fdtable;
+    lock_fdtable(fdtab);
+

     for (size_t i = 0; i < VFS_MAX_FD; i++) {
     for (size_t i = 0; i < VFS_MAX_FD; i++) {

-        if (!__current->fdtable->fds[i]) {
-            *fd = i;
-            return 0;
+        if (__current->fdtable->fds[i]) {
+            continue;

         }
         }

+
+        *fd = i;
+        unlock_fdtable(fdtab);
+        return 0;

     }
     }

+
+    unlock_fdtable(fdtab);

     return EMFILE;
 }
 
     return EMFILE;
 }
 


@@ -348,30 +470,64 @@ vfs_sb_alloc()
     struct v_superblock* sb = cake_grab(superblock_pile);
     memset(sb, 0, sizeof(*sb));
     llist_init_head(&sb->sb_list);
     struct v_superblock* sb = cake_grab(superblock_pile);
     memset(sb, 0, sizeof(*sb));
     llist_init_head(&sb->sb_list);

-    sb->i_cache = vzalloc(VFS_HASHTABLE_SIZE * sizeof(struct hbucket));
+
+    vfs_vncache_init(&sb->i_cache);
+    vfs_vncache_init(&sb->d_cache);
+
+    sb->ref_count = 1;

     return sb;
 }
 
 void
     return sb;
 }
 
 void

-vfs_sb_free(struct v_superblock* sb)
+vfs_sb_ref(struct v_superblock* sb)

 {
 {

-    vfree(sb->i_cache);
+    sb->ref_count++;
+}
+
+void
+vfs_sb_unref(struct v_superblock* sb)
+{
+    assert(sb->ref_count);
+
+    sb->ref_count--;
+    if (likely(sb->ref_count)) {
+        return;
+    }
+
+    if (sb->ops.release) {
+        sb->ops.release(sb);
+    }
+
+    vfs_vncache_free(&sb->i_cache);
+    vfs_vncache_free(&sb->d_cache);
+

     cake_release(superblock_pile, sb);
 }
 
     cake_release(superblock_pile, sb);
 }
 

-static int
+static inline bool
+__dnode_evictable(struct v_dnode* dnode)
+{
+    return dnode->ref_count == 1 
+        && llist_empty(&dnode->children);
+}
+
+static bool

 __vfs_try_evict_dnode(struct lru_node* obj)
 {
     struct v_dnode* dnode = container_of(obj, struct v_dnode, lru);
 
 __vfs_try_evict_dnode(struct lru_node* obj)
 {
     struct v_dnode* dnode = container_of(obj, struct v_dnode, lru);
 

-    if (!dnode->ref_count) {
-        vfs_d_free(dnode);
-        return 1;
+    if (mutex_on_hold(&dnode->lock))
+        return false;
+
+    if (!__dnode_evictable(dnode)) {
+        return false;

     }
     }

-    return 0;
+
+    vfs_d_free(dnode);
+    return true;

 }
 
 }
 

-static int
+static bool

 __vfs_try_evict_inode(struct lru_node* obj)
 {
     struct v_inode* inode = container_of(obj, struct v_inode, lru);
 __vfs_try_evict_inode(struct lru_node* obj)
 {
     struct v_inode* inode = container_of(obj, struct v_inode, lru);


@@ -401,13 +557,12 @@ vfs_d_alloc(struct v_dnode* parent, struct hstr* name)
     llist_init_head(&dnode->aka_list);
     mutex_init(&dnode->lock);
 
     llist_init_head(&dnode->aka_list);
     mutex_init(&dnode->lock);
 

-    dnode->ref_count = ATOMIC_VAR_INIT(0);

     dnode->name = HHSTR(vzalloc(VFS_NAME_MAXLEN), 0, 0);
 
     hstrcpy(&dnode->name, name);
 
     if (parent) {
     dnode->name = HHSTR(vzalloc(VFS_NAME_MAXLEN), 0, 0);
 
     hstrcpy(&dnode->name, name);
 
     if (parent) {

-        dnode->super_block = parent->super_block;
+        vfs_d_assign_sb(dnode, parent->super_block);

         dnode->mnt = parent->mnt;
     }
 
         dnode->mnt = parent->mnt;
     }
 


@@ -420,13 +575,14 @@ void
 vfs_d_free(struct v_dnode* dnode)
 {
     assert(dnode->ref_count == 1);
 vfs_d_free(struct v_dnode* dnode)
 {
     assert(dnode->ref_count == 1);

-
+    

     if (dnode->inode) {
         assert(dnode->inode->link_count > 0);
         dnode->inode->link_count--;
     }
 
     vfs_dcache_remove(dnode);
     if (dnode->inode) {
         assert(dnode->inode->link_count > 0);
         dnode->inode->link_count--;
     }
 
     vfs_dcache_remove(dnode);

+

     // Make sure the children de-referencing their parent.
     // With lru presented, the eviction will be propagated over the entire
     // detached subtree eventually
     // Make sure the children de-referencing their parent.
     // With lru presented, the eviction will be propagated over the entire
     // detached subtree eventually


@@ -436,33 +592,45 @@ vfs_d_free(struct v_dnode* dnode)
         vfs_dcache_remove(pos);
     }
 
         vfs_dcache_remove(pos);
     }
 

-    vfree(dnode->name.value);
+    if (dnode->destruct) {
+        dnode->destruct(dnode);
+    }
+
+    vfs_sb_unref(dnode->super_block);
+    
+    vfree((void*)dnode->name.value);

     cake_release(dnode_pile, dnode);
 }
 
 struct v_inode*
 vfs_i_find(struct v_superblock* sb, u32_t i_id)
 {
     cake_release(dnode_pile, dnode);
 }
 
 struct v_inode*
 vfs_i_find(struct v_superblock* sb, u32_t i_id)
 {

-    struct hbucket* slot = &sb->i_cache[i_id & VFS_HASH_MASK];
-    struct v_inode *pos, *n;
-    hashtable_bucket_foreach(slot, pos, n, hash_list)
+    struct hbucket* slot;
+    struct v_inode *pos, *n, *found = NULL;
+
+    cache_atomic_read(&sb->i_cache, 

     {
     {

-        if (pos->id == i_id) {
+        slot = &sb->i_cache.pool[i_id & VFS_HASH_MASK];
+
+        hashtable_bucket_foreach(slot, pos, n, hash_list)
+        {
+            if (pos->id != i_id) {
+                continue;
+            }
+

             lru_use_one(inode_lru, &pos->lru);
             lru_use_one(inode_lru, &pos->lru);

-            return pos;
+            found = pos;
+            break;

         }
         }

-    }
+    });

 
 

-    return NULL;
+    return found;

 }
 
 void
 vfs_i_addhash(struct v_inode* inode)
 {
 }
 
 void
 vfs_i_addhash(struct v_inode* inode)
 {

-    struct hbucket* slot = &inode->sb->i_cache[inode->id & VFS_HASH_MASK];
-
-    hlist_delete(&inode->hash_list);
-    hlist_add(&slot->head, &inode->hash_list);
+    vfs_vncache_add(inode_cache(inode), inode->id, &inode->hash_list);

 }
 
 struct v_inode*
 }
 
 struct v_inode*


@@ -485,13 +653,13 @@ vfs_i_alloc(struct v_superblock* sb)
 
     sb->ops.init_inode(sb, inode);
 
 
     sb->ops.init_inode(sb, inode);
 

-    inode->sb = sb;

     inode->ctime = clock_unixtime();
     inode->atime = inode->ctime;
     inode->mtime = inode->ctime;
 
     inode->ctime = clock_unixtime();
     inode->atime = inode->ctime;
     inode->mtime = inode->ctime;
 

-done:
+    vfs_i_assign_sb(inode, sb);

     lru_use_one(inode_lru, &inode->lru);
     lru_use_one(inode_lru, &inode->lru);

+    

     return inode;
 }
 
     return inode;
 }
 


@@ -502,104 +670,248 @@ vfs_i_free(struct v_inode* inode)
         pcache_release(inode->pg_cache);
         vfree(inode->pg_cache);
     }
         pcache_release(inode->pg_cache);
         vfree(inode->pg_cache);
     }

+

     // we don't need to sync inode.
     // If an inode can be free, then it must be properly closed.
     // Hence it must be synced already!
     if (inode->destruct) {
         inode->destruct(inode);
     }
     // we don't need to sync inode.
     // If an inode can be free, then it must be properly closed.
     // Hence it must be synced already!
     if (inode->destruct) {
         inode->destruct(inode);
     }

+
+    vfs_sb_unref(inode->sb);
+    

     hlist_delete(&inode->hash_list);
     hlist_delete(&inode->hash_list);

+    lru_remove(inode_lru, &inode->lru);
+

     cake_release(inode_pile, inode);
 }
 
 /* ---- System call definition and support ---- */
 
     cake_release(inode_pile, inode);
 }
 
 /* ---- System call definition and support ---- */
 

-#define FLOCATE_CREATE_EMPTY 1
-#define FLOCATE_CREATE_ONLY 2
+// make a new name when not exists
+#define FLOC_MAYBE_MKNAME 1
+
+// name must be non-exist and made.
+#define FLOC_MKNAME 2
+
+// no follow symlink
+#define FLOC_NOFOLLOW 4

 
 int
 vfs_getfd(int fd, struct v_fd** fd_s)
 {
 
 int
 vfs_getfd(int fd, struct v_fd** fd_s)
 {

-    if (TEST_FD(fd) && (*fd_s = __current->fdtable->fds[fd])) {
-        return 0;
+    struct v_fdtable* fdtab;
+
+    if (!TEST_FD(fd)) {
+        return EBADF;

     }
     }

-    return EBADF;
+
+    fdtab = __current->fdtable;
+
+    lock_fdtable(fdtab);
+    *fd_s = __current->fdtable->fds[fd];
+    unlock_fdtable(fdtab);
+
+    return !*fd_s ? EBADF : 0;

 }
 
 }
 

-int
+static int
+__vfs_mknod(struct v_inode* parent, struct v_dnode* dnode, 
+            unsigned int itype, dev_t* dev)
+{
+    int errno;
+
+    errno = parent->ops->create(parent, dnode, itype);
+    if (errno) {
+        return errno;
+    }
+
+    return 0;
+}
+
+struct file_locator {
+    struct v_dnode* dir;
+    struct v_dnode* file;
+    bool fresh;
+};
+
+/**
+ * @brief unlock the file locator (floc) if possible.
+ *        If the file to be located if not exists, and
+ *        any FLOC_*MKNAME flag is set, then the parent
+ *        dnode will be locked until the file has been properly
+ *        finalised by subsequent logic.
+ * 
+ * @param floc 
+ */
+static inline void
+__floc_try_unlock(struct file_locator* floc)
+{
+    if (floc->fresh) {
+        assert(floc->dir);
+        unlock_dnode(floc->dir);
+    }
+}
+
+static int

 __vfs_try_locate_file(const char* path,
 __vfs_try_locate_file(const char* path,

-                      struct v_dnode** fdir,
-                      struct v_dnode** file,
+                      struct file_locator* floc,

                       int options)
 {
     char name_str[VFS_NAME_MAXLEN];
                       int options)
 {
     char name_str[VFS_NAME_MAXLEN];

+    struct v_dnode *fdir, *file;

     struct hstr name = HSTR(name_str, 0);
     struct hstr name = HSTR(name_str, 0);

-    int errno;
+    int errno, woption = 0;

 
 

+    if ((options & FLOC_NOFOLLOW)) {
+        woption |= VFS_WALK_NOFOLLOW;
+        options &= ~FLOC_NOFOLLOW;
+    }
+
+    floc->fresh = false;

     name_str[0] = 0;
     name_str[0] = 0;

-    if ((errno = vfs_walk_proc(path, fdir, &name, VFS_WALK_PARENT))) {
+    errno = vfs_walk_proc(path, &fdir, &name, woption | VFS_WALK_PARENT);
+    if (errno) {

         return errno;
     }
 
         return errno;
     }
 

-    errno = vfs_walk(*fdir, name.value, file, NULL, 0);
+    lock_dnode(fdir);

 
 

-    if (errno != ENOENT && (options & FLOCATE_CREATE_ONLY)) {
-        return EEXIST;
+    errno = vfs_walk(fdir, name.value, &file, NULL, woption);
+
+    if (errno && errno != ENOENT) {
+        goto error;

     }
 
     }
 

-    if (errno != ENOENT ||
-        !(options & (FLOCATE_CREATE_EMPTY | FLOCATE_CREATE_ONLY))) {
-        return errno;
+    if (!errno && (options & FLOC_MKNAME)) {
+        errno = EEXIST;
+        goto error;
+    }
+    
+    if (!errno) {
+        // the file present, no need to hold the directory lock
+        unlock_dnode(fdir);
+        goto done;

     }
 
     }
 

-    struct v_dnode* parent = *fdir;
-    struct v_dnode* file_new = vfs_d_alloc(parent, &name);
+    // errno == ENOENT
+    if (!options) {
+        goto error;
+    }

 
 

-    if (!file_new) {
-        return ENOMEM;
+    errno = vfs_check_writable(fdir);
+    if (errno) {
+        goto error;

     }
 
     }
 

-    lock_dnode(parent);
+    floc->fresh = true;

 
 

-    if (!(errno = parent->inode->ops->create(parent->inode, file_new))) {
-        vfs_dcache_add(parent, file_new);
-        *file = file_new;
-    } else {
-        vfs_d_free(file_new);
+    file = vfs_d_alloc(fdir, &name);
+
+    if (!file) {
+        errno = ENOMEM;
+        goto error;

     }
 
     }
 

-    unlock_dnode(parent);
+    vfs_dcache_add(fdir, file);
+
+done:
+    floc->dir   = fdir;
+    floc->file  = file;
+    
+    return errno;

 
 

+error:
+    unlock_dnode(fdir);

     return errno;
 }
 
     return errno;
 }
 

+
+static bool
+__check_unlinkable(struct v_dnode* dnode)
+{
+    int acl;
+    bool wr_self, wr_parent;
+    struct v_dnode* parent;
+
+    parent = dnode->parent;
+    acl = dnode->inode->acl;
+
+    wr_self = check_allow_write(dnode->inode);
+    wr_parent = check_allow_write(parent->inode);
+
+    if (!fsacl_test(acl, svtx)) {
+        return wr_self;
+    }
+
+    if (current_euid() == dnode->inode->uid) {
+        return true;
+    }
+
+    return wr_self && wr_parent;
+}
+

 int
 vfs_do_open(const char* path, int options)
 {
 int
 vfs_do_open(const char* path, int options)
 {

-    int errno, fd;
+    int errno, fd, loptions = 0;

     struct v_dnode *dentry, *file;
     struct v_file* ofile = NULL;
     struct v_dnode *dentry, *file;
     struct v_file* ofile = NULL;

+    struct file_locator floc;
+    struct v_inode* inode;
+
+    if ((options & FO_CREATE)) {
+        loptions |= FLOC_MAYBE_MKNAME;
+    } else if ((options & FO_NOFOLLOW)) {
+        loptions |= FLOC_NOFOLLOW;
+    }

 
 

-    errno = __vfs_try_locate_file(
-      path, &dentry, &file, (options & FO_CREATE) ? FLOCATE_CREATE_EMPTY : 0);
+    errno = __vfs_try_locate_file(path, &floc, loptions);

 
 

-    if (!errno && !(errno = vfs_alloc_fdslot(&fd))) {
+    if (errno || (errno = vfs_alloc_fdslot(&fd))) {
+        return errno;
+    }

 
 

-        if (errno || (errno = vfs_open(file, &ofile))) {
+    file   = floc.file;
+    dentry = floc.dir;
+
+    if (floc.fresh) {
+        errno = __vfs_mknod(dentry->inode, file, VFS_IFFILE, NULL);
+        if (errno) {
+            vfs_d_free(file);
+            __floc_try_unlock(&floc);

             return errno;
         }
 
             return errno;
         }
 

-        struct v_fd* fd_s = cake_grab(fd_pile);
-        memset(fd_s, 0, sizeof(*fd_s));
+        __floc_try_unlock(&floc);
+    }

 
 

-        ofile->f_pos = ofile->inode->fsize & -((options & FO_APPEND) != 0);
-        fd_s->file = ofile;
-        fd_s->flags = options;
-        __current->fdtable->fds[fd] = fd_s;
-        return fd;
+
+    if ((errno = vfs_open(file, &ofile))) {
+        return errno;

     }
 
     }
 

-    return errno;
+    inode = ofile->inode;
+    lock_inode(inode);
+
+    struct v_fd* fd_s = cake_grab(fd_pile);
+    memset(fd_s, 0, sizeof(*fd_s));
+
+    if ((options & O_TRUNC)) {
+        file->inode->fsize = 0;   
+    }
+
+    if (vfs_get_dtype(inode->itype) == DT_DIR) {
+        ofile->f_pos = 0;
+    }
+    
+    fd_s->file = ofile;
+    fd_s->flags = options;
+    __current->fdtable->fds[fd] = fd_s;
+
+    unlock_inode(inode);
+    
+    return fd;

 }
 
 __DEFINE_LXSYSCALL2(int, open, const char*, path, int, options)
 }
 
 __DEFINE_LXSYSCALL2(int, open, const char*, path, int, options)


@@ -634,7 +946,7 @@ __vfs_readdir_callback(struct dir_context* dctx,
                        const int dtype)
 {
     struct lx_dirent* dent = (struct lx_dirent*)dctx->cb_data;
                        const int dtype)
 {
     struct lx_dirent* dent = (struct lx_dirent*)dctx->cb_data;

-    strncpy(dent->d_name, name, DIRENT_NAME_MAX_LEN);
+    strncpy(dent->d_name, name, MIN(len, DIRENT_NAME_MAX_LEN));

     dent->d_nlen = len;
     dent->d_type = dtype;
 }
     dent->d_nlen = len;
     dent->d_type = dtype;
 }


@@ -652,29 +964,29 @@ __DEFINE_LXSYSCALL2(int, sys_readdir, int, fd, struct lx_dirent*, dent)
 
     lock_inode(inode);
 
 
     lock_inode(inode);
 

-    if (!(inode->itype & VFS_IFDIR)) {
+    if (!check_directory_node(inode)) {

         errno = ENOTDIR;
         errno = ENOTDIR;

-    } else {
-        struct dir_context dctx =
-          (struct dir_context){ .cb_data = dent,
-                                .index = dent->d_offset,
-                                .read_complete_callback =
-                                  __vfs_readdir_callback };
-        errno = 1;
-        if (dent->d_offset == 0) {
-            __vfs_readdir_callback(&dctx, vfs_dot.value, vfs_dot.len, DT_DIR);
-        } else if (dent->d_offset == 1) {
-            __vfs_readdir_callback(&dctx, vfs_ddot.value, vfs_ddot.len, DT_DIR);
-        } else {
-            dctx.index -= 2;
-            if ((errno = fd_s->file->ops->readdir(fd_s->file, &dctx)) != 1) {
-                unlock_inode(inode);
-                goto done;
-            }
-        }
-        dent->d_offset++;
+        goto unlock;

     }
 
     }
 

+    if (!check_allow_read(inode)) {
+        errno = EPERM;
+        goto unlock;
+    }
+
+    struct dir_context dctx = (struct dir_context) {
+        .cb_data = dent,
+        .read_complete_callback = __vfs_readdir_callback
+    };
+
+    if ((errno = fd_s->file->ops->readdir(fd_s->file, &dctx)) != 1) {
+        goto unlock;
+    }
+
+    dent->d_offset++;
+    fd_s->file->f_pos++;
+
+unlock:

     unlock_inode(inode);
 
 done:
     unlock_inode(inode);
 
 done:


@@ -685,33 +997,41 @@ __DEFINE_LXSYSCALL3(int, read, int, fd, void*, buf, size_t, count)
 {
     int errno = 0;
     struct v_fd* fd_s;
 {
     int errno = 0;
     struct v_fd* fd_s;

+    struct v_inode* inode;
+

     if ((errno = vfs_getfd(fd, &fd_s))) {
         goto done;
     }
 
     struct v_file* file = fd_s->file;
     if ((errno = vfs_getfd(fd, &fd_s))) {
         goto done;
     }
 
     struct v_file* file = fd_s->file;

-    if ((file->inode->itype & VFS_IFDIR)) {
+    if (check_directory_node(file->inode)) {

         errno = EISDIR;
         goto done;
     }
 
         errno = EISDIR;
         goto done;
     }
 

-    lock_inode(file->inode);
+    if (!check_allow_read(file->inode)) {
+        errno = EPERM;
+        goto done;
+    }
+
+    inode = file->inode;
+    lock_inode(inode);

 
 

-    file->inode->atime = clock_unixtime();
+    __vfs_touch_inode(inode, INODE_ACCESSED);

 
 

-    if ((file->inode->itype & VFS_IFSEQDEV) || (fd_s->flags & FO_DIRECT)) {
-        errno = file->ops->read(file->inode, buf, count, file->f_pos);
+    if (check_seqdev_node(inode) || (fd_s->flags & FO_DIRECT)) {
+        errno = file->ops->read(inode, buf, count, file->f_pos);

     } else {
     } else {

-        errno = pcache_read(file->inode, buf, count, file->f_pos);
+        errno = pcache_read(inode, buf, count, file->f_pos);

     }
 
     if (errno > 0) {
         file->f_pos += errno;
     }
 
     if (errno > 0) {
         file->f_pos += errno;

-        unlock_inode(file->inode);
+        unlock_inode(inode);

         return errno;
     }
 
         return errno;
     }
 

-    unlock_inode(file->inode);
+    unlock_inode(inode);

 
 done:
     return DO_STATUS(errno);
 
 done:
     return DO_STATUS(errno);


@@ -725,34 +1045,41 @@ __DEFINE_LXSYSCALL3(int, write, int, fd, void*, buf, size_t, count)
         goto done;
     }
 
         goto done;
     }
 

+    struct v_inode* inode;

     struct v_file* file = fd_s->file;
 
     if ((errno = vfs_check_writable(file->dnode))) {
         goto done;
     }
 
     struct v_file* file = fd_s->file;
 
     if ((errno = vfs_check_writable(file->dnode))) {
         goto done;
     }
 

-    if ((file->inode->itype & VFS_IFDIR)) {
+    if (check_directory_node(file->inode)) {

         errno = EISDIR;
         goto done;
     }
 
         errno = EISDIR;
         goto done;
     }
 

-    lock_inode(file->inode);
+    inode = file->inode;
+    lock_inode(inode);

 
 

-    file->inode->mtime = clock_unixtime();
+    __vfs_touch_inode(inode, INODE_MODIFY);
+    if ((fd_s->flags & O_APPEND)) {
+        file->f_pos = inode->fsize;
+    }

 
 

-    if ((file->inode->itype & VFS_IFSEQDEV) || (fd_s->flags & FO_DIRECT)) {
-        errno = file->ops->write(file->inode, buf, count, file->f_pos);
+    if (check_seqdev_node(inode) || (fd_s->flags & FO_DIRECT)) {
+        errno = file->ops->write(inode, buf, count, file->f_pos);

     } else {
     } else {

-        errno = pcache_write(file->inode, buf, count, file->f_pos);
+        errno = pcache_write(inode, buf, count, file->f_pos);

     }
 
     if (errno > 0) {
         file->f_pos += errno;
     }
 
     if (errno > 0) {
         file->f_pos += errno;

-        unlock_inode(file->inode);
+        inode->fsize = MAX(inode->fsize, file->f_pos);
+
+        unlock_inode(inode);

         return errno;
     }
 
         return errno;
     }
 

-    unlock_inode(file->inode);
+    unlock_inode(inode);

 
 done:
     return DO_STATUS(errno);
 
 done:
     return DO_STATUS(errno);


@@ -767,35 +1094,47 @@ __DEFINE_LXSYSCALL3(int, lseek, int, fd, int, offset, int, options)
     }
 
     struct v_file* file = fd_s->file;
     }
 
     struct v_file* file = fd_s->file;

+    struct v_inode* inode = file->inode;

 
     if (!file->ops->seek) {
         errno = ENOTSUP;
         goto done;
     }
 
 
     if (!file->ops->seek) {
         errno = ENOTSUP;
         goto done;
     }
 

-    lock_inode(file->inode);
+    if (!check_allow_read(inode)) {
+        errno = EPERM;
+        goto done;
+    }
+
+    lock_inode(inode);

 
     int overflow = 0;
     int fpos = file->f_pos;
 
     int overflow = 0;
     int fpos = file->f_pos;

+
+    if (vfs_get_dtype(inode->itype) == DT_DIR) {
+        options = (options != FSEEK_END) ? options : FSEEK_SET;
+    }
+    

     switch (options) {
         case FSEEK_CUR:
     switch (options) {
         case FSEEK_CUR:

-            overflow = __builtin_sadd_overflow((int)file->f_pos, offset, &fpos);
+            overflow = sadd_of((int)file->f_pos, offset, &fpos);

             break;
         case FSEEK_END:
             break;
         case FSEEK_END:

-            overflow =
-              __builtin_sadd_overflow((int)file->inode->fsize, offset, &fpos);
+            overflow = sadd_of((int)inode->fsize, offset, &fpos);

             break;
         case FSEEK_SET:
             fpos = offset;
             break;
     }
             break;
         case FSEEK_SET:
             fpos = offset;
             break;
     }

+

     if (overflow) {
         errno = EOVERFLOW;
     if (overflow) {
         errno = EOVERFLOW;

-    } else if (!(errno = file->ops->seek(file->inode, fpos))) {
-        file->f_pos = fpos;
+    }
+    else {
+        errno = file->ops->seek(file, fpos);

     }
 
     }
 

-    unlock_inode(file->inode);
+    unlock_inode(inode);

 
 done:
     return DO_STATUS(errno);
 
 done:
     return DO_STATUS(errno);


@@ -838,29 +1177,82 @@ vfs_readlink(struct v_dnode* dnode, char* buf, size_t size)
 {
     const char* link;
     struct v_inode* inode = dnode->inode;
 {
     const char* link;
     struct v_inode* inode = dnode->inode;

-    if (inode->ops->read_symlink) {
-        lock_inode(inode);

 
 

-        int errno = inode->ops->read_symlink(inode, &link);
-        strncpy(buf, link, size);
+    if (!check_symlink_node(inode)) {
+        return EINVAL;
+    }

 
 

-        unlock_inode(inode);
-        return errno;
+    if (!inode->ops->read_symlink) {
+        return ENOTSUP;

     }
     }

-    return 0;
+
+    if (!check_allow_read(inode)) {
+        return EPERM;
+    }
+
+    lock_inode(inode);
+
+    int errno = inode->ops->read_symlink(inode, &link);
+    if (errno >= 0) {
+        strncpy(buf, link, MIN(size, (size_t)errno));
+    }
+
+    unlock_inode(inode);
+    return errno;

 }
 
 int
 vfs_get_dtype(int itype)
 {
 }
 
 int
 vfs_get_dtype(int itype)
 {

-    switch (itype) {
-        case VFS_IFDIR:
-            return DT_DIR;
-        case VFS_IFSYMLINK:
-            return DT_SYMLINK;
-        default:
-            return DT_PIPE;
+    int dtype = DT_FILE;
+    if (check_itype(itype, VFS_IFSYMLINK)) {
+        dtype |= DT_SYMLINK;
+    }
+    
+    if (check_itype(itype, VFS_IFDIR)) {
+        dtype |= DT_DIR;
+        return dtype;
+    }
+
+    // TODO other types
+    
+    return dtype;
+}
+
+struct v_fdtable*
+fdtable_create()
+{
+    struct v_fdtable* fdtab;
+
+    fdtab = vzalloc(sizeof(struct v_fdtable));
+    mutex_init(&fdtab->lock);
+
+    return fdtab;
+}
+
+void
+fdtable_copy(struct v_fdtable* dest, struct v_fdtable* src)
+{
+    lock_fdtable(dest);
+    lock_fdtable(src);
+    
+    for (size_t i = 0; i < VFS_MAX_FD; i++) {
+        struct v_fd* fd = src->fds[i];
+        if (!fd)
+            continue;
+        vfs_dup_fd(fd, &dest->fds[i]);

     }
     }

+
+    unlock_fdtable(dest);
+    unlock_fdtable(src);
+}
+
+void
+fdtable_free(struct v_fdtable* table)
+{
+    assert(!mutex_on_hold(&table->lock));
+
+    vfree(table);

 }
 
 __DEFINE_LXSYSCALL3(int, realpathat, int, fd, char*, buf, size_t, size)
 }
 
 __DEFINE_LXSYSCALL3(int, realpathat, int, fd, char*, buf, size_t, size)


@@ -872,11 +1264,12 @@ __DEFINE_LXSYSCALL3(int, realpathat, int, fd, char*, buf, size_t, size)
     }
 
     struct v_dnode* dnode;
     }
 
     struct v_dnode* dnode;

-    errno = vfs_get_path(fd_s->file->dnode, buf, size, 0);

 
 

-    if (errno >= 0) {
-        return errno;
-    }
+    dnode = fd_s->file->dnode;
+
+    lock_dnode(dnode);
+    errno = vfs_get_path(dnode, buf, size, 0);
+    unlock_dnode(dnode);

 
 done:
     return DO_STATUS(errno);
 
 done:
     return DO_STATUS(errno);


@@ -897,16 +1290,8 @@ __DEFINE_LXSYSCALL3(int, readlink, const char*, path, char*, buf, size_t, size)
     return DO_STATUS(errno);
 }
 
     return DO_STATUS(errno);
 }
 

-__DEFINE_LXSYSCALL4(int,
-                    readlinkat,
-                    int,
-                    dirfd,
-                    const char*,
-                    pathname,
-                    char*,
-                    buf,
-                    size_t,
-                    size)
+__DEFINE_LXSYSCALL4(
+  int, readlinkat, int, dirfd, const char*, pathname, char*, buf, size_t, size)

 {
     int errno;
     struct v_fd* fd_s;
 {
     int errno;
     struct v_fd* fd_s;


@@ -914,6 +1299,8 @@ __DEFINE_LXSYSCALL4(int,
         goto done;
     }
 
         goto done;
     }
 

+    pathname = pathname ? pathname : "";
+

     struct v_dnode* dnode;
     if (!(errno = vfs_walk(
             fd_s->file->dnode, pathname, &dnode, NULL, VFS_WALK_NOFOLLOW))) {
     struct v_dnode* dnode;
     if (!(errno = vfs_walk(
             fd_s->file->dnode, pathname, &dnode, NULL, VFS_WALK_NOFOLLOW))) {


@@ -946,6 +1333,11 @@ __DEFINE_LXSYSCALL1(int, rmdir, const char*, pathname)
 
     lock_dnode(dnode);
 
 
     lock_dnode(dnode);
 

+    if (!__check_unlinkable(dnode)) {
+        errno = EPERM;
+        goto done;
+    } 
+

     if ((errno = vfs_check_writable(dnode))) {
         goto done;
     }
     if ((errno = vfs_check_writable(dnode))) {
         goto done;
     }


@@ -975,7 +1367,7 @@ __DEFINE_LXSYSCALL1(int, rmdir, const char*, pathname)
     lock_dnode(parent);
     lock_inode(parent->inode);
 
     lock_dnode(parent);
     lock_inode(parent->inode);
 

-    if ((dnode->inode->itype & VFS_IFDIR)) {
+    if (check_directory_node(dnode->inode)) {

         errno = parent->inode->ops->rmdir(parent->inode, dnode);
         if (!errno) {
             vfs_dcache_remove(dnode);
         errno = parent->inode->ops->rmdir(parent->inode, dnode);
         if (!errno) {
             vfs_dcache_remove(dnode);


@@ -994,15 +1386,23 @@ done:
 
 __DEFINE_LXSYSCALL1(int, mkdir, const char*, path)
 {
 
 __DEFINE_LXSYSCALL1(int, mkdir, const char*, path)
 {

-    int errno = 0;
+    int errno;
+    struct hstr name;
+    struct v_inode* inode;

     struct v_dnode *parent, *dir;
     char name_value[VFS_NAME_MAXLEN];
     struct v_dnode *parent, *dir;
     char name_value[VFS_NAME_MAXLEN];

-    struct hstr name = HHSTR(name_value, 0, 0);
+
+    name = HHSTR(name_value, 0, 0);

 
     if ((errno = vfs_walk_proc(path, &parent, &name, VFS_WALK_PARENT))) {
         goto done;
     }
 
 
     if ((errno = vfs_walk_proc(path, &parent, &name, VFS_WALK_PARENT))) {
         goto done;
     }
 

+    if (!(errno = vfs_walk(parent, name_value, &dir, NULL, 0))) {
+        errno = EEXIST;
+        goto done;
+    }
+

     if ((errno = vfs_check_writable(parent))) {
         goto done;
     }
     if ((errno = vfs_check_writable(parent))) {
         goto done;
     }


@@ -1012,16 +1412,18 @@ __DEFINE_LXSYSCALL1(int, mkdir, const char*, path)
         goto done;
     }
 
         goto done;
     }
 

+    inode = parent->inode;
+

     lock_dnode(parent);
     lock_dnode(parent);

-    lock_inode(parent->inode);
+    lock_inode(inode);

 
     if ((parent->super_block->fs->types & FSTYPE_ROFS)) {
         errno = ENOTSUP;
 
     if ((parent->super_block->fs->types & FSTYPE_ROFS)) {
         errno = ENOTSUP;

-    } else if (!parent->inode->ops->mkdir) {
+    } else if (!inode->ops->mkdir) {

         errno = ENOTSUP;
         errno = ENOTSUP;

-    } else if (!(parent->inode->itype & VFS_IFDIR)) {
+    } else if (!check_directory_node(inode)) {

         errno = ENOTDIR;
         errno = ENOTDIR;

-    } else if (!(errno = parent->inode->ops->mkdir(parent->inode, dir))) {
+    } else if (!(errno = inode->ops->mkdir(inode, dir))) {

         vfs_dcache_add(parent, dir);
         goto cleanup;
     }
         vfs_dcache_add(parent, dir);
         goto cleanup;
     }


@@ -1029,13 +1431,13 @@ __DEFINE_LXSYSCALL1(int, mkdir, const char*, path)
     vfs_d_free(dir);
 
 cleanup:
     vfs_d_free(dir);
 
 cleanup:

-    unlock_inode(parent->inode);
+    unlock_inode(inode);

     unlock_dnode(parent);
 done:
     return DO_STATUS(errno);
 }
 
     unlock_dnode(parent);
 done:
     return DO_STATUS(errno);
 }
 

-int
+static int

 __vfs_do_unlink(struct v_dnode* dnode)
 {
     int errno;
 __vfs_do_unlink(struct v_dnode* dnode)
 {
     int errno;


@@ -1045,6 +1447,10 @@ __vfs_do_unlink(struct v_dnode* dnode)
         return EBUSY;
     }
 
         return EBUSY;
     }
 

+    if (!__check_unlinkable(dnode)) {
+        return EPERM;
+    } 
+

     if ((errno = vfs_check_writable(dnode))) {
         return errno;
     }
     if ((errno = vfs_check_writable(dnode))) {
         return errno;
     }


@@ -1053,10 +1459,8 @@ __vfs_do_unlink(struct v_dnode* dnode)
 
     if (inode->open_count) {
         errno = EBUSY;
 
     if (inode->open_count) {
         errno = EBUSY;

-    } else if (!(inode->itype & VFS_IFDIR)) {
-        // The underlying unlink implementation should handle
-        //  symlink case
-        errno = inode->ops->unlink(inode);
+    } else if (!check_directory_node(inode)) {
+        errno = inode->ops->unlink(inode, dnode);

         if (!errno) {
             vfs_d_free(dnode);
         }
         if (!errno) {
             vfs_d_free(dnode);
         }


@@ -1103,18 +1507,30 @@ done:
 __DEFINE_LXSYSCALL2(int, link, const char*, oldpath, const char*, newpath)
 {
     int errno;
 __DEFINE_LXSYSCALL2(int, link, const char*, oldpath, const char*, newpath)
 {
     int errno;

-    struct v_dnode *dentry, *to_link, *name_dentry, *name_file;
+    struct file_locator floc;
+    struct v_dnode *to_link, *name_file;
+
+    errno = __vfs_try_locate_file(oldpath, &floc, 0);
+    if (errno) {
+        goto done;
+    }

 
 

-    errno = __vfs_try_locate_file(oldpath, &dentry, &to_link, 0);
+    __floc_try_unlock(&floc);
+
+    to_link = floc.file;
+    errno = __vfs_try_locate_file(newpath, &floc, FLOC_MKNAME);

     if (!errno) {
     if (!errno) {

-        errno = __vfs_try_locate_file(
-          newpath, &name_dentry, &name_file, FLOCATE_CREATE_EMPTY);
-        if (!errno) {
-            errno = EEXIST;
-        } else if (name_file) {
-            errno = vfs_link(to_link, name_file);
-        }
+        goto done;       

     }
     }

+
+    name_file = floc.file;
+    errno = vfs_link(to_link, name_file);
+    if (errno) {
+        vfs_d_free(name_file);
+    }
+
+done:
+    __floc_try_unlock(&floc);

     return DO_STATUS(errno);
 }
 
     return DO_STATUS(errno);
 }
 


@@ -1138,7 +1554,7 @@ vfs_dup_fd(struct v_fd* old, struct v_fd** new)
 
     memcpy(copied, old, sizeof(struct v_fd));
 
 
     memcpy(copied, old, sizeof(struct v_fd));
 

-    atomic_fetch_add(&old->file->ref_count, 1);
+    vfs_ref_file(old->file);

 
     *new = copied;
 
 
     *new = copied;
 


@@ -1148,12 +1564,14 @@ vfs_dup_fd(struct v_fd* old, struct v_fd** new)
 int
 vfs_dup2(int oldfd, int newfd)
 {
 int
 vfs_dup2(int oldfd, int newfd)
 {

+    int errno;
+    struct v_fdtable* fdtab;
+    struct v_fd *oldfd_s, *newfd_s;
+    

     if (newfd == oldfd) {
         return newfd;
     }
 
     if (newfd == oldfd) {
         return newfd;
     }
 

-    int errno;
-    struct v_fd *oldfd_s, *newfd_s;

     if ((errno = vfs_getfd(oldfd, &oldfd_s))) {
         goto done;
     }
     if ((errno = vfs_getfd(oldfd, &oldfd_s))) {
         goto done;
     }


@@ -1163,16 +1581,26 @@ vfs_dup2(int oldfd, int newfd)
         goto done;
     }
 
         goto done;
     }
 

-    newfd_s = __current->fdtable->fds[newfd];
+    fdtab = __current->fdtable;
+    lock_fdtable(fdtab);
+
+    newfd_s = fdtab->fds[newfd];

     if (newfd_s && (errno = vfs_close(newfd_s->file))) {
     if (newfd_s && (errno = vfs_close(newfd_s->file))) {

-        goto done;
+        goto unlock_and_done;

     }
 
     }
 

-    if (!(errno = vfs_dup_fd(oldfd_s, &newfd_s))) {
-        __current->fdtable->fds[newfd] = newfd_s;
-        return newfd;
+    if ((errno = vfs_dup_fd(oldfd_s, &newfd_s))) {
+        goto unlock_and_done;

     }
 
     }
 

+    fdtab->fds[newfd] = newfd_s;
+    
+    unlock_fdtable(fdtab);
+    return newfd;
+
+unlock_and_done:
+    unlock_fdtable(fdtab);
+

 done:
     return DO_STATUS(errno);
 }
 done:
     return DO_STATUS(errno);
 }


@@ -1200,81 +1628,79 @@ done:
     return DO_STATUS(errno);
 }
 
     return DO_STATUS(errno);
 }
 

-__DEFINE_LXSYSCALL2(int,
-                    symlink,
-                    const char*,
-                    pathname,
-                    const char*,
-                    link_target)
+__DEFINE_LXSYSCALL2(
+  int, symlink, const char*, pathname, const char*, link_target)

 {
     int errno;
 {
     int errno;

-    struct v_dnode *dnode, *file;
-    if ((errno = __vfs_try_locate_file(
-           pathname, &dnode, &file, FLOCATE_CREATE_ONLY))) {
+    struct file_locator floc;
+    struct v_dnode *file;
+    struct v_inode *f_ino;
+    
+    errno = __vfs_try_locate_file(pathname, &floc, FLOC_MKNAME);
+    if (errno) {

         goto done;
     }
 
         goto done;
     }
 

-    if (errno = vfs_check_writable(file)) {
+    file = floc.file;
+    errno = __vfs_mknod(floc.dir->inode, file, VFS_IFSYMLINK, NULL);
+    if (errno) {
+        vfs_d_free(file);

         goto done;
     }
 
         goto done;
     }
 

-    if (!file->inode->ops->set_symlink) {
+    f_ino = file->inode;
+
+    assert(f_ino);
+
+    errno = vfs_check_writable(file);
+    if (errno) {
+        goto done;
+    }
+
+    if (!f_ino->ops->set_symlink) {

         errno = ENOTSUP;
         goto done;
     }
 
         errno = ENOTSUP;
         goto done;
     }
 

-    lock_inode(file->inode);
+    lock_inode(f_ino);

 
 

-    errno = file->inode->ops->set_symlink(file->inode, link_target);
+    errno = f_ino->ops->set_symlink(f_ino, link_target);

 
 

-    unlock_inode(file->inode);
+    unlock_inode(f_ino);

 
 done:
 
 done:

+    __floc_try_unlock(&floc);

     return DO_STATUS(errno);
 }
 
     return DO_STATUS(errno);
 }
 

-void
-vfs_ref_file(struct v_file* file)
+static int
+vfs_do_chdir_nolock(struct proc_info* proc, struct v_dnode* dnode)

 {
 {

-    atomic_fetch_add(&file->ref_count, 1);
-}
+    if (!check_directory_node(dnode->inode)) {
+        return ENOTDIR;
+    }

 
 

-void
-vfs_ref_dnode(struct v_dnode* dnode)
-{
-    atomic_fetch_add(&dnode->ref_count, 1);
-    mnt_mkbusy(dnode->mnt);
-}
+    if (proc->cwd) {
+        vfs_unref_dnode(proc->cwd);
+    }

 
 

-void
-vfs_unref_dnode(struct v_dnode* dnode)
-{
-    atomic_fetch_sub(&dnode->ref_count, 1);
-    mnt_chillax(dnode->mnt);
+    vfs_ref_dnode(dnode);
+    proc->cwd = dnode;
+
+    return 0;

 }
 
 }
 

-int
+static int

 vfs_do_chdir(struct proc_info* proc, struct v_dnode* dnode)
 {
     int errno = 0;
 
     lock_dnode(dnode);
 
 vfs_do_chdir(struct proc_info* proc, struct v_dnode* dnode)
 {
     int errno = 0;
 
     lock_dnode(dnode);
 

-    if (!(dnode->inode->itype & VFS_IFDIR)) {
-        errno = ENOTDIR;
-        goto done;
-    }
-
-    if (proc->cwd) {
-        vfs_unref_dnode(proc->cwd);
-    }
-
-    vfs_ref_dnode(dnode);
-    proc->cwd = dnode;
+    errno = vfs_do_chdir_nolock(proc, dnode);

 
     unlock_dnode(dnode);
 
 
     unlock_dnode(dnode);
 

-done:

     return errno;
 }
 
     return errno;
 }
 


@@ -1287,7 +1713,7 @@ __DEFINE_LXSYSCALL1(int, chdir, const char*, path)
         goto done;
     }
 
         goto done;
     }
 

-    errno = vfs_do_chdir(__current, dnode);
+    errno = vfs_do_chdir((struct proc_info*)__current, dnode);

 
 done:
     return DO_STATUS(errno);
 
 done:
     return DO_STATUS(errno);


@@ -1302,7 +1728,32 @@ __DEFINE_LXSYSCALL1(int, fchdir, int, fd)
         goto done;
     }
 
         goto done;
     }
 

-    errno = vfs_do_chdir(__current, fd_s->file->dnode);
+    errno = vfs_do_chdir((struct proc_info*)__current, fd_s->file->dnode);
+
+done:
+    return DO_STATUS(errno);
+}
+
+
+__DEFINE_LXSYSCALL1(int, chroot, const char*, path)
+{
+    int errno;
+    struct v_dnode* dnode;
+    if ((errno = vfs_walk_proc(path, &dnode, NULL, 0))) {
+        return errno;
+    }
+
+    lock_dnode(dnode);
+
+    errno = vfs_do_chdir_nolock(__current, dnode);
+    if (errno) {
+        unlock_dnode(dnode);
+        goto done;
+    }
+
+    __current->root = dnode;
+    
+    unlock_dnode(dnode);

 
 done:
     return DO_STATUS(errno);
 
 done:
     return DO_STATUS(errno);


@@ -1335,7 +1786,7 @@ __DEFINE_LXSYSCALL2(char*, getcwd, char*, buf, size_t, size)
     ret_ptr = buf;
 
 done:
     ret_ptr = buf;
 
 done:

-    __current->k_status = errno;
+    syscall_result(errno);

     return ret_ptr;
 }
 
     return ret_ptr;
 }
 


@@ -1348,7 +1799,7 @@ vfs_do_rename(struct v_dnode* current, struct v_dnode* target)
         return 0;
     }
 
         return 0;
     }
 

-    if (errno = vfs_check_writable(current)) {
+    if ((errno = vfs_check_writable(current))) {

         return errno;
     }
 
         return errno;
     }
 


@@ -1365,6 +1816,7 @@ vfs_do_rename(struct v_dnode* current, struct v_dnode* target)
 
     lock_dnode(current);
     lock_dnode(target);
 
     lock_dnode(current);
     lock_dnode(target);

+

     if (oldparent)
         lock_dnode(oldparent);
     if (newparent)
     if (oldparent)
         lock_dnode(oldparent);
     if (newparent)


@@ -1393,6 +1845,7 @@ vfs_do_rename(struct v_dnode* current, struct v_dnode* target)
 
 cleanup:
     unlock_dnode(current);
 
 cleanup:
     unlock_dnode(current);

+    

     if (oldparent)
         unlock_dnode(oldparent);
     if (newparent)
     if (oldparent)
         unlock_dnode(oldparent);
     if (newparent)


@@ -1432,6 +1885,158 @@ __DEFINE_LXSYSCALL2(int, rename, const char*, oldpath, const char*, newpath)
     errno = vfs_do_rename(cur, target);
 
 done:
     errno = vfs_do_rename(cur, target);
 
 done:

-    vfree(name.value);
+    vfree((void*)name.value);
+    return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL2(int, fstat, int, fd, struct file_stat*, stat)
+{
+    int errno = 0;
+    struct v_fd* fds;
+
+    if ((errno = vfs_getfd(fd, &fds))) {
+        goto done;
+    }
+
+    struct v_inode* vino = fds->file->inode;
+    struct device* fdev = vino->sb->dev;
+
+    stat->st_ino     = vino->id;
+    stat->st_blocks  = vino->lb_usage;
+    stat->st_size    = vino->fsize;
+    stat->st_blksize = vino->sb->blksize;
+    stat->st_nlink   = vino->link_count;
+    stat->st_uid     = vino->uid;
+    stat->st_gid     = vino->gid;
+
+    stat->st_ctim    = vino->ctime;
+    stat->st_atim    = vino->atime;
+    stat->st_mtim    = vino->mtime;
+
+    stat->st_mode    = (vino->itype << 16) | vino->acl;
+
+    stat->st_ioblksize = PAGE_SIZE;
+
+    if (check_device_node(vino)) {
+        struct device* rdev = resolve_device(vino->data);
+        if (!rdev) {
+            errno = EINVAL;
+            goto done;
+        }
+
+        stat->st_rdev = (dev_t){.meta = rdev->ident.fn_grp,
+                                .unique = rdev->ident.unique,
+                                .index = dev_uid(rdev) };
+    }
+
+    if (fdev) {
+        stat->st_dev = (dev_t){.meta = fdev->ident.fn_grp,
+                               .unique = fdev->ident.unique,
+                               .index = dev_uid(fdev) };
+    }
+
+done:
+    return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL4(int, fchmodat, int, fd, 
+                    const char*, path, int, mode, int, flags)
+{
+    int errno;
+    struct v_dnode *dnode;
+    struct v_inode* inode;
+
+    errno = vfs_walkat(fd, path, flags, &dnode);
+    if (errno) {
+        goto done;
+    }
+
+    errno = vfs_check_writable(dnode);
+    if (errno) {
+        return errno;
+    }
+
+    inode = dnode->inode;
+    lock_inode(inode);
+
+    if (!current_is_root()) {
+        mode = mode & FSACL_RWXMASK;
+    }
+
+    inode->acl = mode;
+    __vfs_touch_inode(inode, INODE_MODIFY);
+
+    unlock_inode(inode);
+
+done:
+    return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL5(int, fchownat, int, fd, 
+                    const char*, path, uid_t, uid, gid_t, gid, int, flags)
+{
+    int errno;
+    struct v_dnode *dnode;
+    struct v_inode *inode;
+
+    errno = vfs_walkat(fd, path, flags, &dnode);
+    if (errno) {
+        goto done;
+    }
+
+    errno = vfs_check_writable(dnode);
+    if (errno) {
+        return errno;
+    }
+
+    inode = dnode->inode;
+    lock_inode(inode);
+
+    inode->uid = uid;
+    inode->gid = gid;
+    __vfs_touch_inode(inode, INODE_MODIFY);
+
+    unlock_inode(inode);
+
+done:
+    return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL4(int, faccessat, int, fd, 
+                    const char*, path, int, amode, int, flags)
+{
+    int errno, acl;
+    struct v_dnode *dnode;
+    struct v_inode *inode;
+    struct user_scope* uscope;
+
+    uid_t tuid;
+    gid_t tgid;
+
+    errno = vfs_walkat(fd, path, flags, &dnode);
+    if (errno) {
+        goto done;
+    }
+
+    if ((flags & AT_EACCESS)) {
+        tuid = current_euid();
+        tgid = current_egid();
+    }
+    else {
+        uscope = current_user_scope();
+        tuid = uscope->ruid;
+        tgid = uscope->rgid;
+    }
+
+    inode = dnode->inode;
+
+    acl  = inode->acl;
+    acl &= amode;
+    acl &= check_acl_between(inode->uid, inode->gid, tuid, tgid);
+    if (!acl) {
+        errno = EACCESS;
+    }
+
+done:

     return DO_STATUS(errno);
 }
\ No newline at end of file
     return DO_STATUS(errno);
 }
\ No newline at end of file