mirror_zfs/module/zfs/zpl_inode.c

808 lines
18 KiB
C
Raw Normal View History

/*
* 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 (c) 2011, Lawrence Livermore National Security, LLC.
* Copyright (c) 2015 by Chunwei Chen. All rights reserved.
*/
Linux 3.18 compat: Snapshot auto-mounting Re-factor the .zfs/snapshot auto-mouting code to take in to account changes made to the upstream kernels. And to lay the groundwork for enabling access to .zfs snapshots via NFS clients. This patch makes the following core improvements. * All actively auto-mounted snapshots are now tracked in two global trees which are indexed by snapshot name and objset id respectively. This allows for fast lookups of any auto-mounted snapshot regardless without needing access to the parent dataset. * Snapshot entries are added to the tree in zfsctl_snapshot_mount(). However, they are now removed from the tree in the context of the unmount process. This eliminates the need complicated error logic in zfsctl_snapshot_unmount() to handle unmount failures. * References are now taken on the snapshot entries in the tree to ensure they always remain valid while a task is outstanding. * The MNT_SHRINKABLE flag is set on the snapshot vfsmount_t right after the auto-mount succeeds. This allows to kernel to unmount idle auto-mounted snapshots if needed removing the need for the zfsctl_unmount_snapshots() function. * Snapshots in active use will not be automatically unmounted. As long as at least one dentry is revalidated every zfs_expire_snapshot/2 seconds the auto-unmount expiration timer will be extended. * Commit torvalds/linux@bafc9b7 caused snapshots auto-mounted by ZFS to be immediately unmounted when the dentry was revalidated. This was a consequence of ZFS invaliding all snapdir dentries to ensure that negative dentries didn't mask new snapshots. This patch modifies the behavior such that only negative dentries are invalidated. This solves the issue and may result in a performance improvement. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #3589 Closes #3344 Closes #3295 Closes #3257 Closes #3243 Closes #3030 Closes #2841
2015-04-25 02:21:13 +03:00
#include <sys/zfs_ctldir.h>
#include <sys/zfs_vfsops.h>
#include <sys/zfs_vnops.h>
#include <sys/zfs_znode.h>
#include <sys/dmu_objset.h>
#include <sys/vfs.h>
#include <sys/zpl.h>
#include <sys/file.h>
static struct dentry *
#ifdef HAVE_LOOKUP_NAMEIDATA
zpl_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd)
#else
zpl_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags)
#endif
{
cred_t *cr = CRED();
struct inode *ip;
int error;
fstrans_cookie_t cookie;
pathname_t *ppn = NULL;
pathname_t pn;
int zfs_flags = 0;
zfs_sb_t *zsb = dentry->d_sb->s_fs_info;
if (dlen(dentry) > ZFS_MAX_DATASET_NAME_LEN)
return (ERR_PTR(-ENAMETOOLONG));
crhold(cr);
cookie = spl_fstrans_mark();
/* If we are a case insensitive fs, we need the real name */
if (zsb->z_case == ZFS_CASE_INSENSITIVE) {
zfs_flags = FIGNORECASE;
pn_alloc(&pn);
ppn = &pn;
}
error = -zfs_lookup(dir, dname(dentry), &ip, zfs_flags, cr, NULL, ppn);
spl_fstrans_unmark(cookie);
ASSERT3S(error, <=, 0);
crfree(cr);
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
spin_lock(&dentry->d_lock);
dentry->d_time = jiffies;
#ifndef HAVE_S_D_OP
d_set_d_op(dentry, &zpl_dentry_operations);
#endif /* HAVE_S_D_OP */
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
spin_unlock(&dentry->d_lock);
if (error) {
/*
* If we have a case sensitive fs, we do not want to
* insert negative entries, so return NULL for ENOENT.
* Fall through if the error is not ENOENT. Also free memory.
*/
if (ppn) {
pn_free(ppn);
if (error == -ENOENT)
return (NULL);
}
if (error == -ENOENT)
return (d_splice_alias(NULL, dentry));
else
return (ERR_PTR(error));
}
/*
* If we are case insensitive, call the correct function
* to install the name.
*/
if (ppn) {
struct dentry *new_dentry;
struct qstr ci_name;
if (strcmp(dname(dentry), pn.pn_buf) == 0) {
new_dentry = d_splice_alias(ip, dentry);
} else {
ci_name.name = pn.pn_buf;
ci_name.len = strlen(pn.pn_buf);
new_dentry = d_add_ci(dentry, ip, &ci_name);
}
pn_free(ppn);
return (new_dentry);
} else {
return (d_splice_alias(ip, dentry));
}
}
void
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
zpl_vap_init(vattr_t *vap, struct inode *dir, zpl_umode_t mode, cred_t *cr)
{
vap->va_mask = ATTR_MODE;
vap->va_mode = mode;
vap->va_uid = crgetfsuid(cr);
if (dir && dir->i_mode & S_ISGID) {
vap->va_gid = KGID_TO_SGID(dir->i_gid);
if (S_ISDIR(mode))
vap->va_mode |= S_ISGID;
} else {
vap->va_gid = crgetfsgid(cr);
}
}
static int
#ifdef HAVE_CREATE_NAMEIDATA
zpl_create(struct inode *dir, struct dentry *dentry, zpl_umode_t mode,
struct nameidata *nd)
#else
zpl_create(struct inode *dir, struct dentry *dentry, zpl_umode_t mode,
bool flag)
#endif
{
cred_t *cr = CRED();
struct inode *ip;
vattr_t *vap;
int error;
fstrans_cookie_t cookie;
crhold(cr);
vap = kmem_zalloc(sizeof (vattr_t), KM_SLEEP);
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
zpl_vap_init(vap, dir, mode, cr);
cookie = spl_fstrans_mark();
error = -zfs_create(dir, dname(dentry), vap, 0, mode, &ip, cr, 0, NULL);
if (error == 0) {
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
d_instantiate(dentry, ip);
error = zpl_xattr_security_init(ip, dir, &dentry->d_name);
if (error == 0)
error = zpl_init_acl(ip, dir);
if (error)
(void) zfs_remove(dir, dname(dentry), cr, 0);
}
spl_fstrans_unmark(cookie);
kmem_free(vap, sizeof (vattr_t));
crfree(cr);
ASSERT3S(error, <=, 0);
return (error);
}
static int
zpl_mknod(struct inode *dir, struct dentry *dentry, zpl_umode_t mode,
dev_t rdev)
{
cred_t *cr = CRED();
struct inode *ip;
vattr_t *vap;
int error;
fstrans_cookie_t cookie;
/*
* We currently expect Linux to supply rdev=0 for all sockets
* and fifos, but we want to know if this behavior ever changes.
*/
if (S_ISSOCK(mode) || S_ISFIFO(mode))
ASSERT(rdev == 0);
crhold(cr);
vap = kmem_zalloc(sizeof (vattr_t), KM_SLEEP);
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
zpl_vap_init(vap, dir, mode, cr);
vap->va_rdev = rdev;
cookie = spl_fstrans_mark();
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
error = -zfs_create(dir, dname(dentry), vap, 0, mode, &ip, cr, 0, NULL);
if (error == 0) {
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
d_instantiate(dentry, ip);
error = zpl_xattr_security_init(ip, dir, &dentry->d_name);
if (error == 0)
error = zpl_init_acl(ip, dir);
if (error)
(void) zfs_remove(dir, dname(dentry), cr, 0);
}
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
spl_fstrans_unmark(cookie);
kmem_free(vap, sizeof (vattr_t));
crfree(cr);
ASSERT3S(error, <=, 0);
return (error);
}
#ifdef HAVE_TMPFILE
static int
zpl_tmpfile(struct inode *dir, struct dentry *dentry, zpl_umode_t mode)
{
cred_t *cr = CRED();
struct inode *ip;
vattr_t *vap;
int error;
fstrans_cookie_t cookie;
crhold(cr);
vap = kmem_zalloc(sizeof (vattr_t), KM_SLEEP);
zpl_vap_init(vap, dir, mode, cr);
cookie = spl_fstrans_mark();
error = -zfs_tmpfile(dir, vap, 0, mode, &ip, cr, 0, NULL);
if (error == 0) {
/* d_tmpfile will do drop_nlink, so we should set it first */
set_nlink(ip, 1);
d_tmpfile(dentry, ip);
error = zpl_xattr_security_init(ip, dir, &dentry->d_name);
if (error == 0)
error = zpl_init_acl(ip, dir);
/*
* don't need to handle error here, file is already in
* unlinked set.
*/
}
spl_fstrans_unmark(cookie);
kmem_free(vap, sizeof (vattr_t));
crfree(cr);
ASSERT3S(error, <=, 0);
return (error);
}
#endif
static int
zpl_unlink(struct inode *dir, struct dentry *dentry)
{
cred_t *cr = CRED();
int error;
fstrans_cookie_t cookie;
zfs_sb_t *zsb = dentry->d_sb->s_fs_info;
crhold(cr);
cookie = spl_fstrans_mark();
error = -zfs_remove(dir, dname(dentry), cr, 0);
/*
* For a CI FS we must invalidate the dentry to prevent the
* creation of negative entries.
*/
if (error == 0 && zsb->z_case == ZFS_CASE_INSENSITIVE)
d_invalidate(dentry);
spl_fstrans_unmark(cookie);
crfree(cr);
ASSERT3S(error, <=, 0);
return (error);
}
static int
zpl_mkdir(struct inode *dir, struct dentry *dentry, zpl_umode_t mode)
{
cred_t *cr = CRED();
vattr_t *vap;
struct inode *ip;
int error;
fstrans_cookie_t cookie;
crhold(cr);
vap = kmem_zalloc(sizeof (vattr_t), KM_SLEEP);
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
zpl_vap_init(vap, dir, mode | S_IFDIR, cr);
cookie = spl_fstrans_mark();
error = -zfs_mkdir(dir, dname(dentry), vap, &ip, cr, 0, NULL);
if (error == 0) {
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
d_instantiate(dentry, ip);
error = zpl_xattr_security_init(ip, dir, &dentry->d_name);
if (error == 0)
error = zpl_init_acl(ip, dir);
if (error)
(void) zfs_rmdir(dir, dname(dentry), NULL, cr, 0);
}
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
spl_fstrans_unmark(cookie);
kmem_free(vap, sizeof (vattr_t));
crfree(cr);
ASSERT3S(error, <=, 0);
return (error);
}
static int
zpl_rmdir(struct inode * dir, struct dentry *dentry)
{
cred_t *cr = CRED();
int error;
fstrans_cookie_t cookie;
zfs_sb_t *zsb = dentry->d_sb->s_fs_info;
crhold(cr);
cookie = spl_fstrans_mark();
error = -zfs_rmdir(dir, dname(dentry), NULL, cr, 0);
/*
* For a CI FS we must invalidate the dentry to prevent the
* creation of negative entries.
*/
if (error == 0 && zsb->z_case == ZFS_CASE_INSENSITIVE)
d_invalidate(dentry);
spl_fstrans_unmark(cookie);
crfree(cr);
ASSERT3S(error, <=, 0);
return (error);
}
static int
zpl_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
{
int error;
fstrans_cookie_t cookie;
cookie = spl_fstrans_mark();
Improve fstat(2) performance There is at most a factor of 3x performance improvement to be had by using the Linux generic_fillattr() helper. However, to use it safely we need to ensure the values in a cached inode are kept rigerously up to date. Unfortunately, this isn't the case for the blksize, blocks, and atime fields. At the moment the authoritative values are still stored in the znode. This patch introduces an optimized zfs_getattr_fast() call. The idea is to use the up to date values from the inode and the blksize, block, and atime fields from the znode. At some latter date we should be able to strictly use the inode values and further improve performance. The remaining overhead in the zfs_getattr_fast() call can be attributed to having to take the znode mutex. This overhead is unavoidable until the inode is kept strictly up to date. The the careful reader will notice the we do not use the customary ZFS_ENTER()/ZFS_EXIT() macros. These macro's are designed to ensure the filesystem is not torn down in the middle of an operation. However, in this case the VFS is holding a reference on the active inode so we know this is impossible. =================== Performance Tests ======================== This test calls the fstat(2) system call 10,000,000 times on an open file description in a tight loop. The test results show the zfs stat(2) performance is now only 22% slower than ext4. This is a 2.5x improvement and there is a clear long term plan to get to parity with ext4. filesystem | test-1 test-2 test-3 | average | times-ext4 --------------+-------------------------+---------+----------- ext4 | 7.785s 7.899s 7.284s | 7.656s | 1.000x zfs-0.6.0-rc4 | 24.052s 22.531s 23.857s | 23.480s | 3.066x zfs-faststat | 9.224s 9.398s 9.485s | 9.369s | 1.223x The second test is to run 'du' of a copy of the /usr tree which contains 110514 files. The test is run multiple times both using both a cold cache (/proc/sys/vm/drop_caches) and a hot cache. As expected this change signigicantly improved the zfs hot cache performance and doesn't quite bring zfs to parity with ext4. A little surprisingly the zfs cold cache performance is better than ext4. This can probably be attributed to the zfs allocation policy of co-locating all the meta data on disk which minimizes seek times. By default the ext4 allocator will spread the data over the entire disk only co-locating each directory. filesystem | cold | hot --------------+---------+-------- ext4 | 13.318s | 1.040s zfs-0.6.0-rc4 | 4.982s | 1.762s zfs-faststat | 4.933s | 1.345s
2011-07-10 02:44:16 +04:00
error = -zfs_getattr_fast(dentry->d_inode, stat);
spl_fstrans_unmark(cookie);
ASSERT3S(error, <=, 0);
return (error);
}
static int
Drop HAVE_XVATTR macros When I began work on the Posix layer it immediately became clear to me that to integrate cleanly with the Linux VFS certain Solaris specific things would have to go. One of these things was to elimate as many Solaris specific types from the ZPL layer as possible. They would be replaced with their Linux equivalents. This would not only be good for performance, but for the general readability and health of the code. The Solaris and Linux VFS are different beasts and should be treated as such. Most of the code remains common for constructing transactions and such, but there are subtle and important differenced which need to be repsected. This policy went quite for for certain types such as the vnode_t, and it initially seemed to be working out well for the vattr_t. There was a relatively small amount of related xvattr_t code I was forced to comment out with HAVE_XVATTR. But it didn't look that hard to come back soon and replace it all with a native Linux type. However, after going doing this path with xvattr some distance it clear that this code was woven in the ZPL more deeply than I thought. In particular its hooks went very deep in to the ZPL replay code and replacing it would not be as easy as I originally thought. Rather than continue persuing replacing and removing this code I've taken a step back and reevaluted things. This commit reverts many of my previous commits which removed xvattr related code. It restores much of the code to its original upstream state and now relies on improved xvattr_t support in the zfs package itself. The result of this is that much of the code which I had commented out, which accidentally broke things like replay, is now back in place and working. However, there may be a small performance impact for getattr/setattr operations because they now require a translation from native Linux to Solaris types. For now that's a price I'm willing to pay. Once everything is completely functional we can revisting the issue of removing the vattr_t/xvattr_t types. Closes #111
2011-03-01 23:24:09 +03:00
zpl_setattr(struct dentry *dentry, struct iattr *ia)
{
struct inode *ip = dentry->d_inode;
cred_t *cr = CRED();
Drop HAVE_XVATTR macros When I began work on the Posix layer it immediately became clear to me that to integrate cleanly with the Linux VFS certain Solaris specific things would have to go. One of these things was to elimate as many Solaris specific types from the ZPL layer as possible. They would be replaced with their Linux equivalents. This would not only be good for performance, but for the general readability and health of the code. The Solaris and Linux VFS are different beasts and should be treated as such. Most of the code remains common for constructing transactions and such, but there are subtle and important differenced which need to be repsected. This policy went quite for for certain types such as the vnode_t, and it initially seemed to be working out well for the vattr_t. There was a relatively small amount of related xvattr_t code I was forced to comment out with HAVE_XVATTR. But it didn't look that hard to come back soon and replace it all with a native Linux type. However, after going doing this path with xvattr some distance it clear that this code was woven in the ZPL more deeply than I thought. In particular its hooks went very deep in to the ZPL replay code and replacing it would not be as easy as I originally thought. Rather than continue persuing replacing and removing this code I've taken a step back and reevaluted things. This commit reverts many of my previous commits which removed xvattr related code. It restores much of the code to its original upstream state and now relies on improved xvattr_t support in the zfs package itself. The result of this is that much of the code which I had commented out, which accidentally broke things like replay, is now back in place and working. However, there may be a small performance impact for getattr/setattr operations because they now require a translation from native Linux to Solaris types. For now that's a price I'm willing to pay. Once everything is completely functional we can revisting the issue of removing the vattr_t/xvattr_t types. Closes #111
2011-03-01 23:24:09 +03:00
vattr_t *vap;
int error;
fstrans_cookie_t cookie;
error = setattr_prepare(dentry, ia);
if (error)
return (error);
crhold(cr);
vap = kmem_zalloc(sizeof (vattr_t), KM_SLEEP);
Drop HAVE_XVATTR macros When I began work on the Posix layer it immediately became clear to me that to integrate cleanly with the Linux VFS certain Solaris specific things would have to go. One of these things was to elimate as many Solaris specific types from the ZPL layer as possible. They would be replaced with their Linux equivalents. This would not only be good for performance, but for the general readability and health of the code. The Solaris and Linux VFS are different beasts and should be treated as such. Most of the code remains common for constructing transactions and such, but there are subtle and important differenced which need to be repsected. This policy went quite for for certain types such as the vnode_t, and it initially seemed to be working out well for the vattr_t. There was a relatively small amount of related xvattr_t code I was forced to comment out with HAVE_XVATTR. But it didn't look that hard to come back soon and replace it all with a native Linux type. However, after going doing this path with xvattr some distance it clear that this code was woven in the ZPL more deeply than I thought. In particular its hooks went very deep in to the ZPL replay code and replacing it would not be as easy as I originally thought. Rather than continue persuing replacing and removing this code I've taken a step back and reevaluted things. This commit reverts many of my previous commits which removed xvattr related code. It restores much of the code to its original upstream state and now relies on improved xvattr_t support in the zfs package itself. The result of this is that much of the code which I had commented out, which accidentally broke things like replay, is now back in place and working. However, there may be a small performance impact for getattr/setattr operations because they now require a translation from native Linux to Solaris types. For now that's a price I'm willing to pay. Once everything is completely functional we can revisting the issue of removing the vattr_t/xvattr_t types. Closes #111
2011-03-01 23:24:09 +03:00
vap->va_mask = ia->ia_valid & ATTR_IATTR_MASK;
vap->va_mode = ia->ia_mode;
vap->va_uid = KUID_TO_SUID(ia->ia_uid);
vap->va_gid = KGID_TO_SGID(ia->ia_gid);
Drop HAVE_XVATTR macros When I began work on the Posix layer it immediately became clear to me that to integrate cleanly with the Linux VFS certain Solaris specific things would have to go. One of these things was to elimate as many Solaris specific types from the ZPL layer as possible. They would be replaced with their Linux equivalents. This would not only be good for performance, but for the general readability and health of the code. The Solaris and Linux VFS are different beasts and should be treated as such. Most of the code remains common for constructing transactions and such, but there are subtle and important differenced which need to be repsected. This policy went quite for for certain types such as the vnode_t, and it initially seemed to be working out well for the vattr_t. There was a relatively small amount of related xvattr_t code I was forced to comment out with HAVE_XVATTR. But it didn't look that hard to come back soon and replace it all with a native Linux type. However, after going doing this path with xvattr some distance it clear that this code was woven in the ZPL more deeply than I thought. In particular its hooks went very deep in to the ZPL replay code and replacing it would not be as easy as I originally thought. Rather than continue persuing replacing and removing this code I've taken a step back and reevaluted things. This commit reverts many of my previous commits which removed xvattr related code. It restores much of the code to its original upstream state and now relies on improved xvattr_t support in the zfs package itself. The result of this is that much of the code which I had commented out, which accidentally broke things like replay, is now back in place and working. However, there may be a small performance impact for getattr/setattr operations because they now require a translation from native Linux to Solaris types. For now that's a price I'm willing to pay. Once everything is completely functional we can revisting the issue of removing the vattr_t/xvattr_t types. Closes #111
2011-03-01 23:24:09 +03:00
vap->va_size = ia->ia_size;
vap->va_atime = ia->ia_atime;
vap->va_mtime = ia->ia_mtime;
vap->va_ctime = ia->ia_ctime;
if (vap->va_mask & ATTR_ATIME)
ip->i_atime = timespec_trunc(ia->ia_atime,
ip->i_sb->s_time_gran);
cookie = spl_fstrans_mark();
error = -zfs_setattr(ip, vap, 0, cr);
if (!error && (ia->ia_valid & ATTR_MODE))
error = zpl_chmod_acl(ip);
Drop HAVE_XVATTR macros When I began work on the Posix layer it immediately became clear to me that to integrate cleanly with the Linux VFS certain Solaris specific things would have to go. One of these things was to elimate as many Solaris specific types from the ZPL layer as possible. They would be replaced with their Linux equivalents. This would not only be good for performance, but for the general readability and health of the code. The Solaris and Linux VFS are different beasts and should be treated as such. Most of the code remains common for constructing transactions and such, but there are subtle and important differenced which need to be repsected. This policy went quite for for certain types such as the vnode_t, and it initially seemed to be working out well for the vattr_t. There was a relatively small amount of related xvattr_t code I was forced to comment out with HAVE_XVATTR. But it didn't look that hard to come back soon and replace it all with a native Linux type. However, after going doing this path with xvattr some distance it clear that this code was woven in the ZPL more deeply than I thought. In particular its hooks went very deep in to the ZPL replay code and replacing it would not be as easy as I originally thought. Rather than continue persuing replacing and removing this code I've taken a step back and reevaluted things. This commit reverts many of my previous commits which removed xvattr related code. It restores much of the code to its original upstream state and now relies on improved xvattr_t support in the zfs package itself. The result of this is that much of the code which I had commented out, which accidentally broke things like replay, is now back in place and working. However, there may be a small performance impact for getattr/setattr operations because they now require a translation from native Linux to Solaris types. For now that's a price I'm willing to pay. Once everything is completely functional we can revisting the issue of removing the vattr_t/xvattr_t types. Closes #111
2011-03-01 23:24:09 +03:00
spl_fstrans_unmark(cookie);
kmem_free(vap, sizeof (vattr_t));
crfree(cr);
ASSERT3S(error, <=, 0);
Drop HAVE_XVATTR macros When I began work on the Posix layer it immediately became clear to me that to integrate cleanly with the Linux VFS certain Solaris specific things would have to go. One of these things was to elimate as many Solaris specific types from the ZPL layer as possible. They would be replaced with their Linux equivalents. This would not only be good for performance, but for the general readability and health of the code. The Solaris and Linux VFS are different beasts and should be treated as such. Most of the code remains common for constructing transactions and such, but there are subtle and important differenced which need to be repsected. This policy went quite for for certain types such as the vnode_t, and it initially seemed to be working out well for the vattr_t. There was a relatively small amount of related xvattr_t code I was forced to comment out with HAVE_XVATTR. But it didn't look that hard to come back soon and replace it all with a native Linux type. However, after going doing this path with xvattr some distance it clear that this code was woven in the ZPL more deeply than I thought. In particular its hooks went very deep in to the ZPL replay code and replacing it would not be as easy as I originally thought. Rather than continue persuing replacing and removing this code I've taken a step back and reevaluted things. This commit reverts many of my previous commits which removed xvattr related code. It restores much of the code to its original upstream state and now relies on improved xvattr_t support in the zfs package itself. The result of this is that much of the code which I had commented out, which accidentally broke things like replay, is now back in place and working. However, there may be a small performance impact for getattr/setattr operations because they now require a translation from native Linux to Solaris types. For now that's a price I'm willing to pay. Once everything is completely functional we can revisting the issue of removing the vattr_t/xvattr_t types. Closes #111
2011-03-01 23:24:09 +03:00
return (error);
}
static int
zpl_rename2(struct inode *sdip, struct dentry *sdentry,
struct inode *tdip, struct dentry *tdentry, unsigned int flags)
{
cred_t *cr = CRED();
int error;
fstrans_cookie_t cookie;
/* We don't have renameat2(2) support */
if (flags)
return (-EINVAL);
crhold(cr);
cookie = spl_fstrans_mark();
error = -zfs_rename(sdip, dname(sdentry), tdip, dname(tdentry), cr, 0);
spl_fstrans_unmark(cookie);
crfree(cr);
ASSERT3S(error, <=, 0);
return (error);
}
#ifndef HAVE_RENAME_WANTS_FLAGS
static int
zpl_rename(struct inode *sdip, struct dentry *sdentry,
struct inode *tdip, struct dentry *tdentry)
{
return (zpl_rename2(sdip, sdentry, tdip, tdentry, 0));
}
#endif
static int
zpl_symlink(struct inode *dir, struct dentry *dentry, const char *name)
{
cred_t *cr = CRED();
vattr_t *vap;
struct inode *ip;
int error;
fstrans_cookie_t cookie;
crhold(cr);
vap = kmem_zalloc(sizeof (vattr_t), KM_SLEEP);
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
zpl_vap_init(vap, dir, S_IFLNK | S_IRWXUGO, cr);
cookie = spl_fstrans_mark();
error = -zfs_symlink(dir, dname(dentry), vap, (char *)name, &ip, cr, 0);
if (error == 0) {
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
d_instantiate(dentry, ip);
error = zpl_xattr_security_init(ip, dir, &dentry->d_name);
if (error)
(void) zfs_remove(dir, dname(dentry), cr, 0);
}
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
spl_fstrans_unmark(cookie);
kmem_free(vap, sizeof (vattr_t));
crfree(cr);
ASSERT3S(error, <=, 0);
return (error);
}
#if defined(HAVE_PUT_LINK_COOKIE)
static void
zpl_put_link(struct inode *unused, void *cookie)
{
kmem_free(cookie, MAXPATHLEN);
}
#elif defined(HAVE_PUT_LINK_NAMEIDATA)
static void
zpl_put_link(struct dentry *dentry, struct nameidata *nd, void *ptr)
{
const char *link = nd_get_link(nd);
if (!IS_ERR(link))
kmem_free(link, MAXPATHLEN);
}
#elif defined(HAVE_PUT_LINK_DELAYED)
static void
zpl_put_link(void *ptr)
{
kmem_free(ptr, MAXPATHLEN);
}
#endif
static int
zpl_get_link_common(struct dentry *dentry, struct inode *ip, char **link)
{
fstrans_cookie_t cookie;
cred_t *cr = CRED();
struct iovec iov;
uio_t uio;
int error;
crhold(cr);
*link = NULL;
iov.iov_len = MAXPATHLEN;
iov.iov_base = kmem_zalloc(MAXPATHLEN, KM_SLEEP);
uio.uio_iov = &iov;
uio.uio_iovcnt = 1;
uio.uio_skip = 0;
uio.uio_resid = (MAXPATHLEN - 1);
uio.uio_segflg = UIO_SYSSPACE;
cookie = spl_fstrans_mark();
error = -zfs_readlink(ip, &uio, cr);
spl_fstrans_unmark(cookie);
crfree(cr);
if (error)
kmem_free(iov.iov_base, MAXPATHLEN);
else
*link = iov.iov_base;
return (error);
}
#if defined(HAVE_GET_LINK_DELAYED)
const char *
zpl_get_link(struct dentry *dentry, struct inode *inode,
struct delayed_call *done)
{
char *link = NULL;
int error;
if (!dentry)
return (ERR_PTR(-ECHILD));
error = zpl_get_link_common(dentry, inode, &link);
if (error)
return (ERR_PTR(error));
set_delayed_call(done, zpl_put_link, link);
return (link);
}
#elif defined(HAVE_GET_LINK_COOKIE)
const char *
zpl_get_link(struct dentry *dentry, struct inode *inode, void **cookie)
{
char *link = NULL;
int error;
if (!dentry)
return (ERR_PTR(-ECHILD));
error = zpl_get_link_common(dentry, inode, &link);
if (error)
return (ERR_PTR(error));
return (*cookie = link);
}
#elif defined(HAVE_FOLLOW_LINK_COOKIE)
const char *
zpl_follow_link(struct dentry *dentry, void **cookie)
{
char *link = NULL;
int error;
error = zpl_get_link_common(dentry, dentry->d_inode, &link);
if (error)
return (ERR_PTR(error));
return (*cookie = link);
}
#elif defined(HAVE_FOLLOW_LINK_NAMEIDATA)
static void *
zpl_follow_link(struct dentry *dentry, struct nameidata *nd)
{
char *link = NULL;
int error;
error = zpl_get_link_common(dentry, dentry->d_inode, &link);
if (error)
nd_set_link(nd, ERR_PTR(error));
else
nd_set_link(nd, link);
return (NULL);
}
#endif
static int
zpl_link(struct dentry *old_dentry, struct inode *dir, struct dentry *dentry)
{
cred_t *cr = CRED();
struct inode *ip = old_dentry->d_inode;
int error;
fstrans_cookie_t cookie;
if (ip->i_nlink >= ZFS_LINK_MAX)
return (-EMLINK);
crhold(cr);
ip->i_ctime = CURRENT_TIME_SEC;
igrab(ip); /* Use ihold() if available */
cookie = spl_fstrans_mark();
error = -zfs_link(dir, ip, dname(dentry), cr, 0);
if (error) {
iput(ip);
goto out;
}
d_instantiate(dentry, ip);
out:
spl_fstrans_unmark(cookie);
crfree(cr);
ASSERT3S(error, <=, 0);
return (error);
}
#ifdef HAVE_INODE_TRUNCATE_RANGE
static void
zpl_truncate_range(struct inode *ip, loff_t start, loff_t end)
{
cred_t *cr = CRED();
flock64_t bf;
fstrans_cookie_t cookie;
ASSERT3S(start, <=, end);
/*
* zfs_freesp() will interpret (len == 0) as meaning "truncate until
* the end of the file". We don't want that.
*/
if (start == end)
return;
crhold(cr);
bf.l_type = F_WRLCK;
bf.l_whence = 0;
bf.l_start = start;
bf.l_len = end - start;
bf.l_pid = 0;
cookie = spl_fstrans_mark();
zfs_space(ip, F_FREESP, &bf, FWRITE, start, cr);
spl_fstrans_unmark(cookie);
crfree(cr);
}
#endif /* HAVE_INODE_TRUNCATE_RANGE */
#ifdef HAVE_INODE_FALLOCATE
static long
zpl_fallocate(struct inode *ip, int mode, loff_t offset, loff_t len)
{
return (zpl_fallocate_common(ip, mode, offset, len));
}
#endif /* HAVE_INODE_FALLOCATE */
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
static int
#ifdef HAVE_D_REVALIDATE_NAMEIDATA
zpl_revalidate(struct dentry *dentry, struct nameidata *nd)
{
unsigned int flags = (nd ? nd->flags : 0);
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
#else
zpl_revalidate(struct dentry *dentry, unsigned int flags)
{
#endif /* HAVE_D_REVALIDATE_NAMEIDATA */
zfs_sb_t *zsb = dentry->d_sb->s_fs_info;
int error;
if (flags & LOOKUP_RCU)
return (-ECHILD);
Linux 3.18 compat: Snapshot auto-mounting Re-factor the .zfs/snapshot auto-mouting code to take in to account changes made to the upstream kernels. And to lay the groundwork for enabling access to .zfs snapshots via NFS clients. This patch makes the following core improvements. * All actively auto-mounted snapshots are now tracked in two global trees which are indexed by snapshot name and objset id respectively. This allows for fast lookups of any auto-mounted snapshot regardless without needing access to the parent dataset. * Snapshot entries are added to the tree in zfsctl_snapshot_mount(). However, they are now removed from the tree in the context of the unmount process. This eliminates the need complicated error logic in zfsctl_snapshot_unmount() to handle unmount failures. * References are now taken on the snapshot entries in the tree to ensure they always remain valid while a task is outstanding. * The MNT_SHRINKABLE flag is set on the snapshot vfsmount_t right after the auto-mount succeeds. This allows to kernel to unmount idle auto-mounted snapshots if needed removing the need for the zfsctl_unmount_snapshots() function. * Snapshots in active use will not be automatically unmounted. As long as at least one dentry is revalidated every zfs_expire_snapshot/2 seconds the auto-unmount expiration timer will be extended. * Commit torvalds/linux@bafc9b7 caused snapshots auto-mounted by ZFS to be immediately unmounted when the dentry was revalidated. This was a consequence of ZFS invaliding all snapdir dentries to ensure that negative dentries didn't mask new snapshots. This patch modifies the behavior such that only negative dentries are invalidated. This solves the issue and may result in a performance improvement. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #3589 Closes #3344 Closes #3295 Closes #3257 Closes #3243 Closes #3030 Closes #2841
2015-04-25 02:21:13 +03:00
/*
* Automounted snapshots rely on periodic dentry revalidation
* to defer snapshots from being automatically unmounted.
*/
if (zsb->z_issnap) {
if (time_after(jiffies, zsb->z_snap_defer_time +
MAX(zfs_expire_snapshot * HZ / 2, HZ))) {
zsb->z_snap_defer_time = jiffies;
zfsctl_snapshot_unmount_delay(zsb->z_os->os_spa,
Linux 3.18 compat: Snapshot auto-mounting Re-factor the .zfs/snapshot auto-mouting code to take in to account changes made to the upstream kernels. And to lay the groundwork for enabling access to .zfs snapshots via NFS clients. This patch makes the following core improvements. * All actively auto-mounted snapshots are now tracked in two global trees which are indexed by snapshot name and objset id respectively. This allows for fast lookups of any auto-mounted snapshot regardless without needing access to the parent dataset. * Snapshot entries are added to the tree in zfsctl_snapshot_mount(). However, they are now removed from the tree in the context of the unmount process. This eliminates the need complicated error logic in zfsctl_snapshot_unmount() to handle unmount failures. * References are now taken on the snapshot entries in the tree to ensure they always remain valid while a task is outstanding. * The MNT_SHRINKABLE flag is set on the snapshot vfsmount_t right after the auto-mount succeeds. This allows to kernel to unmount idle auto-mounted snapshots if needed removing the need for the zfsctl_unmount_snapshots() function. * Snapshots in active use will not be automatically unmounted. As long as at least one dentry is revalidated every zfs_expire_snapshot/2 seconds the auto-unmount expiration timer will be extended. * Commit torvalds/linux@bafc9b7 caused snapshots auto-mounted by ZFS to be immediately unmounted when the dentry was revalidated. This was a consequence of ZFS invaliding all snapdir dentries to ensure that negative dentries didn't mask new snapshots. This patch modifies the behavior such that only negative dentries are invalidated. This solves the issue and may result in a performance improvement. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #3589 Closes #3344 Closes #3295 Closes #3257 Closes #3243 Closes #3030 Closes #2841
2015-04-25 02:21:13 +03:00
dmu_objset_id(zsb->z_os), zfs_expire_snapshot);
}
}
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
/*
* After a rollback negative dentries created before the rollback
* time must be invalidated. Otherwise they can obscure files which
* are only present in the rolled back dataset.
*/
if (dentry->d_inode == NULL) {
spin_lock(&dentry->d_lock);
error = time_before(dentry->d_time, zsb->z_rollback_time);
spin_unlock(&dentry->d_lock);
if (error)
return (0);
}
/*
* The dentry may reference a stale inode if a mounted file system
* was rolled back to a point in time where the object didn't exist.
*/
if (dentry->d_inode && ITOZ(dentry->d_inode)->z_is_stale)
return (0);
return (1);
}
const struct inode_operations zpl_inode_operations = {
.setattr = zpl_setattr,
.getattr = zpl_getattr,
#ifdef HAVE_GENERIC_SETXATTR
.setxattr = generic_setxattr,
.getxattr = generic_getxattr,
.removexattr = generic_removexattr,
#endif
.listxattr = zpl_xattr_list,
#ifdef HAVE_INODE_TRUNCATE_RANGE
.truncate_range = zpl_truncate_range,
#endif /* HAVE_INODE_TRUNCATE_RANGE */
#ifdef HAVE_INODE_FALLOCATE
.fallocate = zpl_fallocate,
#endif /* HAVE_INODE_FALLOCATE */
#if defined(CONFIG_FS_POSIX_ACL)
#if defined(HAVE_GET_ACL)
.get_acl = zpl_get_acl,
#elif defined(HAVE_CHECK_ACL)
.check_acl = zpl_check_acl,
#elif defined(HAVE_PERMISSION)
.permission = zpl_permission,
#endif /* HAVE_GET_ACL | HAVE_CHECK_ACL | HAVE_PERMISSION */
#endif /* CONFIG_FS_POSIX_ACL */
};
const struct inode_operations zpl_dir_inode_operations = {
.create = zpl_create,
.lookup = zpl_lookup,
.link = zpl_link,
.unlink = zpl_unlink,
.symlink = zpl_symlink,
.mkdir = zpl_mkdir,
.rmdir = zpl_rmdir,
.mknod = zpl_mknod,
#ifdef HAVE_RENAME_WANTS_FLAGS
.rename = zpl_rename2,
#else
.rename = zpl_rename,
#endif
#ifdef HAVE_TMPFILE
.tmpfile = zpl_tmpfile,
#endif
.setattr = zpl_setattr,
.getattr = zpl_getattr,
#ifdef HAVE_GENERIC_SETXATTR
.setxattr = generic_setxattr,
.getxattr = generic_getxattr,
.removexattr = generic_removexattr,
#endif
.listxattr = zpl_xattr_list,
#if defined(CONFIG_FS_POSIX_ACL)
#if defined(HAVE_GET_ACL)
.get_acl = zpl_get_acl,
#elif defined(HAVE_CHECK_ACL)
.check_acl = zpl_check_acl,
#elif defined(HAVE_PERMISSION)
.permission = zpl_permission,
#endif /* HAVE_GET_ACL | HAVE_CHECK_ACL | HAVE_PERMISSION */
#endif /* CONFIG_FS_POSIX_ACL */
};
const struct inode_operations zpl_symlink_inode_operations = {
.readlink = generic_readlink,
#if defined(HAVE_GET_LINK_DELAYED) || defined(HAVE_GET_LINK_COOKIE)
.get_link = zpl_get_link,
#elif defined(HAVE_FOLLOW_LINK_COOKIE) || defined(HAVE_FOLLOW_LINK_NAMEIDATA)
.follow_link = zpl_follow_link,
#endif
#if defined(HAVE_PUT_LINK_COOKIE) || defined(HAVE_PUT_LINK_NAMEIDATA)
.put_link = zpl_put_link,
#endif
.setattr = zpl_setattr,
.getattr = zpl_getattr,
#ifdef HAVE_GENERIC_SETXATTR
.setxattr = generic_setxattr,
.getxattr = generic_getxattr,
.removexattr = generic_removexattr,
#endif
.listxattr = zpl_xattr_list,
};
const struct inode_operations zpl_special_inode_operations = {
.setattr = zpl_setattr,
.getattr = zpl_getattr,
#ifdef HAVE_GENERIC_SETXATTR
.setxattr = generic_setxattr,
.getxattr = generic_getxattr,
.removexattr = generic_removexattr,
#endif
.listxattr = zpl_xattr_list,
#if defined(CONFIG_FS_POSIX_ACL)
#if defined(HAVE_GET_ACL)
.get_acl = zpl_get_acl,
#elif defined(HAVE_CHECK_ACL)
.check_acl = zpl_check_acl,
#elif defined(HAVE_PERMISSION)
.permission = zpl_permission,
#endif /* HAVE_GET_ACL | HAVE_CHECK_ACL | HAVE_PERMISSION */
#endif /* CONFIG_FS_POSIX_ACL */
};
Fix 'zfs rollback' on mounted file systems Rolling back a mounted filesystem with open file handles and cached dentries+inodes never worked properly in ZoL. The major issue was that Linux provides no easy mechanism for modules to invalidate the inode cache for a file system. Because of this it was possible that an inode from the previous filesystem would not get properly dropped from the cache during rolling back. Then a new inode with the same inode number would be create and collide with the existing cached inode. Ideally this would trigger an VERIFY() but in practice the error wasn't handled and it would just NULL reference. Luckily, this issue can be resolved by sprucing up the existing Solaris zfs_rezget() functionality for the Linux VFS. The way it works now is that when a file system is rolled back all the cached inodes will be traversed and refetched from disk. If a version of the cached inode exists on disk the in-core copy will be updated accordingly. If there is no match for that object on disk it will be unhashed from the inode cache and marked as stale. This will effectively make the inode unfindable for lookups allowing the inode number to be immediately recycled. The inode will then only be accessible from the cached dentries. Subsequent dentry lookups which reference a stale inode will result in the dentry being invalidated. Once invalidated the dentry will drop its reference on the inode allowing it to be safely pruned from the cache. Special care is taken for negative dentries since they do not reference any inode. These dentires will be invalidate based on when they were added to the dentry cache. Entries added before the last rollback will be invalidate to prevent them from masking real files in the dataset. Two nice side effects of this fix are: * Removes the dependency on spl_invalidate_inodes(), it can now be safely removed from the SPL when we choose to do so. * zfs_znode_alloc() no longer requires a dentry to be passed. This effectively reverts this portition of the code to its upstream counterpart. The dentry is not instantiated more correctly in the Linux ZPL layer. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ned Bass <bass6@llnl.gov> Closes #795
2013-01-16 04:41:09 +04:00
dentry_operations_t zpl_dentry_operations = {
.d_revalidate = zpl_revalidate,
};