#include <linux/mm.h>
#include <linux/dcache.h>
#include <linux/init.h>
-#include <linux/quotaops.h>
#include <linux/slab.h>
#include <linux/writeback.h>
#include <linux/module.h>
#include <linux/backing-dev.h>
#include <linux/wait.h>
+#include <linux/rwsem.h>
#include <linux/hash.h>
#include <linux/swap.h>
#include <linux/security.h>
-#include <linux/ima.h>
#include <linux/pagemap.h>
#include <linux/cdev.h>
#include <linux/bootmem.h>
DEFINE_SPINLOCK(inode_lock);
/*
- * iprune_mutex provides exclusion between the kswapd or try_to_free_pages
+ * iprune_sem provides exclusion between the kswapd or try_to_free_pages
* icache shrinking path, and the umount path. Without this exclusion,
* by the time prune_icache calls iput for the inode whose pages it has
* been invalidating, or by the time it calls clear_inode & destroy_inode
* from its final dispose_list, the struct super_block they refer to
* (for inode->i_sb->s_op) may already have been freed and reused.
+ *
+ * We make this an rwsem because the fastpath is icache shrinking. In
+ * some cases a filesystem may be doing a significant amount of work in
+ * its inode reclaim code, so this should improve parallelism.
*/
-static DEFINE_MUTEX(iprune_mutex);
+static DECLARE_RWSEM(iprune_sem);
/*
* Statistics gathering..
* Prevent speculative execution through spin_unlock(&inode_lock);
*/
smp_mb();
- wake_up_bit(&inode->i_state, __I_LOCK);
+ wake_up_bit(&inode->i_state, __I_NEW);
}
/**
if (security_inode_alloc(inode))
goto out;
-
- /* allocate and initialize an i_integrity */
- if (ima_inode_alloc(inode))
- goto out_free_security;
-
spin_lock_init(&inode->i_lock);
lockdep_set_class(&inode->i_lock, &sb->s_type->i_lock_key);
#endif
return 0;
-
-out_free_security:
- security_inode_free(inode);
out:
return -ENOMEM;
}
void __destroy_inode(struct inode *inode)
{
BUG_ON(inode_has_buffers(inode));
- ima_inode_free(inode);
security_inode_free(inode);
fsnotify_inode_delete(inode);
#ifdef CONFIG_FS_POSIX_ACL
BUG_ON(!(inode->i_state & I_FREEING));
BUG_ON(inode->i_state & I_CLEAR);
inode_sync_wait(inode);
- vfs_dq_drop(inode);
if (inode->i_sb->s_op->clear_inode)
inode->i_sb->s_op->clear_inode(inode);
if (S_ISBLK(inode->i_mode) && inode->i_bdev)
/*
* We can reschedule here without worrying about the list's
* consistency because the per-sb list of inodes must not
- * change during umount anymore, and because iprune_mutex keeps
+ * change during umount anymore, and because iprune_sem keeps
* shrink_icache_memory() away.
*/
cond_resched_lock(&inode_lock);
int busy;
LIST_HEAD(throw_away);
- mutex_lock(&iprune_mutex);
+ down_write(&iprune_sem);
spin_lock(&inode_lock);
inotify_unmount_inodes(&sb->s_inodes);
fsnotify_unmount_inodes(&sb->s_inodes);
spin_unlock(&inode_lock);
dispose_list(&throw_away);
- mutex_unlock(&iprune_mutex);
+ up_write(&iprune_sem);
return busy;
}
int nr_scanned;
unsigned long reap = 0;
- mutex_lock(&iprune_mutex);
+ down_read(&iprune_sem);
spin_lock(&inode_lock);
for (nr_scanned = 0; nr_scanned < nr_to_scan; nr_scanned++) {
struct inode *inode;
spin_unlock(&inode_lock);
dispose_list(&freeable);
- mutex_unlock(&iprune_mutex);
+ up_read(&iprune_sem);
}
/*
}
#endif
/*
- * This is special! We do not need the spinlock when clearing I_LOCK,
+ * This is special! We do not need the spinlock when clearing I_NEW,
* because we're guaranteed that nobody else tries to do anything about
* the state of the inode when it is locked, as we just created it (so
- * there can be no old holders that haven't tested I_LOCK).
+ * there can be no old holders that haven't tested I_NEW).
* However we must emit the memory barrier so that other CPUs reliably
- * see the clearing of I_LOCK after the other inode initialisation has
+ * see the clearing of I_NEW after the other inode initialisation has
* completed.
*/
smp_mb();
- WARN_ON((inode->i_state & (I_LOCK|I_NEW)) != (I_LOCK|I_NEW));
- inode->i_state &= ~(I_LOCK|I_NEW);
+ WARN_ON(!(inode->i_state & I_NEW));
+ inode->i_state &= ~I_NEW;
wake_up_inode(inode);
}
EXPORT_SYMBOL(unlock_new_inode);
goto set_failed;
__inode_add_to_lists(sb, head, inode);
- inode->i_state = I_LOCK|I_NEW;
+ inode->i_state = I_NEW;
spin_unlock(&inode_lock);
/* Return the locked inode with I_NEW set, the
if (!old) {
inode->i_ino = ino;
__inode_add_to_lists(sb, head, inode);
- inode->i_state = I_LOCK|I_NEW;
+ inode->i_state = I_NEW;
spin_unlock(&inode_lock);
/* Return the locked inode with I_NEW set, the
ino_t ino = inode->i_ino;
struct hlist_head *head = inode_hashtable + hash(sb, ino);
- inode->i_state |= I_LOCK|I_NEW;
+ inode->i_state |= I_NEW;
while (1) {
struct hlist_node *node;
struct inode *old = NULL;
struct super_block *sb = inode->i_sb;
struct hlist_head *head = inode_hashtable + hash(sb, hashval);
- inode->i_state |= I_LOCK|I_NEW;
+ inode->i_state |= I_NEW;
while (1) {
struct hlist_node *node;
if (op->delete_inode) {
void (*delete)(struct inode *) = op->delete_inode;
- if (!is_bad_inode(inode))
- vfs_dq_init(inode);
/* Filesystems implementing their own
* s_op->delete_inode are required to call
* truncate_inode_pages and clear_inode()
}
EXPORT_SYMBOL(generic_delete_inode);
-static void generic_forget_inode(struct inode *inode)
+/**
+ * generic_detach_inode - remove inode from inode lists
+ * @inode: inode to remove
+ *
+ * Remove inode from inode lists, write it if it's dirty. This is just an
+ * internal VFS helper exported for hugetlbfs. Do not use!
+ *
+ * Returns 1 if inode should be completely destroyed.
+ */
+int generic_detach_inode(struct inode *inode)
{
struct super_block *sb = inode->i_sb;
inodes_stat.nr_unused++;
if (sb->s_flags & MS_ACTIVE) {
spin_unlock(&inode_lock);
- return;
+ return 0;
}
WARN_ON(inode->i_state & I_NEW);
inode->i_state |= I_WILL_FREE;
inode->i_state |= I_FREEING;
inodes_stat.nr_inodes--;
spin_unlock(&inode_lock);
+ return 1;
+}
+EXPORT_SYMBOL_GPL(generic_detach_inode);
+
+static void generic_forget_inode(struct inode *inode)
+{
+ if (!generic_detach_inode(inode))
+ return;
if (inode->i_data.nrpages)
truncate_inode_pages(&inode->i_data, 0);
clear_inode(inode);
struct inode *inode = dentry->d_inode;
struct timespec now;
- if (mnt_want_write(mnt))
- return;
if (inode->i_flags & S_NOATIME)
- goto out;
+ return;
if (IS_NOATIME(inode))
- goto out;
+ return;
if ((inode->i_sb->s_flags & MS_NODIRATIME) && S_ISDIR(inode->i_mode))
- goto out;
+ return;
if (mnt->mnt_flags & MNT_NOATIME)
- goto out;
+ return;
if ((mnt->mnt_flags & MNT_NODIRATIME) && S_ISDIR(inode->i_mode))
- goto out;
+ return;
now = current_fs_time(inode->i_sb);
if (!relatime_need_update(mnt, inode, now))
- goto out;
+ return;
if (timespec_equal(&inode->i_atime, &now))
- goto out;
+ return;
+
+ if (mnt_want_write(mnt))
+ return;
inode->i_atime = now;
mark_inode_dirty_sync(inode);
-out:
mnt_drop_write(mnt);
}
EXPORT_SYMBOL(touch_atime);
{
struct inode *inode = file->f_path.dentry->d_inode;
struct timespec now;
- int sync_it = 0;
- int err;
+ enum { S_MTIME = 1, S_CTIME = 2, S_VERSION = 4 } sync_it = 0;
+ /* First try to exhaust all avenues to not sync */
if (IS_NOCMTIME(inode))
return;
- err = mnt_want_write_file(file);
- if (err)
- return;
-
now = current_fs_time(inode->i_sb);
- if (!timespec_equal(&inode->i_mtime, &now)) {
- inode->i_mtime = now;
- sync_it = 1;
- }
+ if (!timespec_equal(&inode->i_mtime, &now))
+ sync_it = S_MTIME;
- if (!timespec_equal(&inode->i_ctime, &now)) {
- inode->i_ctime = now;
- sync_it = 1;
- }
+ if (!timespec_equal(&inode->i_ctime, &now))
+ sync_it |= S_CTIME;
- if (IS_I_VERSION(inode)) {
- inode_inc_iversion(inode);
- sync_it = 1;
- }
+ if (IS_I_VERSION(inode))
+ sync_it |= S_VERSION;
+
+ if (!sync_it)
+ return;
- if (sync_it)
- mark_inode_dirty_sync(inode);
+ /* Finally allowed to write? Takes lock. */
+ if (mnt_want_write_file(file))
+ return;
+
+ /* Only change inode inside the lock region */
+ if (sync_it & S_VERSION)
+ inode_inc_iversion(inode);
+ if (sync_it & S_CTIME)
+ inode->i_ctime = now;
+ if (sync_it & S_MTIME)
+ inode->i_mtime = now;
+ mark_inode_dirty_sync(inode);
mnt_drop_write(file->f_path.mnt);
}
EXPORT_SYMBOL(file_update_time);
* until the deletion _might_ have completed. Callers are responsible
* to recheck inode state.
*
- * It doesn't matter if I_LOCK is not set initially, a call to
+ * It doesn't matter if I_NEW is not set initially, a call to
* wake_up_inode() after removing from the hash list will DTRT.
*
* This is called with inode_lock held.
static void __wait_on_freeing_inode(struct inode *inode)
{
wait_queue_head_t *wq;
- DEFINE_WAIT_BIT(wait, &inode->i_state, __I_LOCK);
- wq = bit_waitqueue(&inode->i_state, __I_LOCK);
+ DEFINE_WAIT_BIT(wait, &inode->i_state, __I_NEW);
+ wq = bit_waitqueue(&inode->i_state, __I_NEW);
prepare_to_wait(wq, &wait.wait, TASK_UNINTERRUPTIBLE);
spin_unlock(&inode_lock);
schedule();
else if (S_ISSOCK(mode))
inode->i_fop = &bad_sock_fops;
else
- printk(KERN_DEBUG "init_special_inode: bogus i_mode (%o)\n",
- mode);
+ printk(KERN_DEBUG "init_special_inode: bogus i_mode (%o) for"
+ " inode %s:%lu\n", mode, inode->i_sb->s_id,
+ inode->i_ino);
}
EXPORT_SYMBOL(init_special_inode);