#include "xfs_trans.h"
#include "xfs_sb.h"
#include "xfs_ag.h"
-#include "xfs_dir.h"
#include "xfs_dir2.h"
#include "xfs_alloc.h"
#include "xfs_dmapi.h"
#include "xfs_bmap_btree.h"
#include "xfs_alloc_btree.h"
#include "xfs_ialloc_btree.h"
-#include "xfs_dir_sf.h"
#include "xfs_dir2_sf.h"
#include "xfs_attr_sf.h"
#include "xfs_dinode.h"
#include "xfs_itable.h"
#include "xfs_rw.h"
#include "xfs_acl.h"
-#include "xfs_cap.h"
-#include "xfs_mac.h"
#include "xfs_attr.h"
#include "xfs_buf_item.h"
#include "xfs_utils.h"
+#include "xfs_vnodeops.h"
+#include <linux/capability.h>
#include <linux/xattr.h>
#include <linux/namei.h>
+#include <linux/security.h>
+#include <linux/falloc.h>
-#define IS_NOATIME(inode) ((inode->i_sb->s_flags & MS_NOATIME) || \
- (S_ISDIR(inode->i_mode) && inode->i_sb->s_flags & MS_NODIRATIME))
+/*
+ * Bring the atime in the XFS inode uptodate.
+ * Used before logging the inode to disk or when the Linux inode goes away.
+ */
+void
+xfs_synchronize_atime(
+ xfs_inode_t *ip)
+{
+ struct inode *inode = VFS_I(ip);
+
+ if (inode) {
+ ip->i_d.di_atime.t_sec = (__int32_t)inode->i_atime.tv_sec;
+ ip->i_d.di_atime.t_nsec = (__int32_t)inode->i_atime.tv_nsec;
+ }
+}
+
+/*
+ * If the linux inode exists, mark it dirty.
+ * Used when commiting a dirty inode into a transaction so that
+ * the inode will get written back by the linux code
+ */
+void
+xfs_mark_inode_dirty_sync(
+ xfs_inode_t *ip)
+{
+ struct inode *inode = VFS_I(ip);
+
+ if (inode)
+ mark_inode_dirty_sync(inode);
+}
/*
* Change the requested timestamp in the given inode.
* We don't lock across timestamp updates, and we don't log them but
* we do record the fact that there is dirty information in core.
- *
- * NOTE -- callers MUST combine XFS_ICHGTIME_MOD or XFS_ICHGTIME_CHG
- * with XFS_ICHGTIME_ACC to be sure that access time
- * update will take. Calling first with XFS_ICHGTIME_ACC
- * and then XFS_ICHGTIME_MOD may fail to modify the access
- * timestamp if the filesystem is mounted noacctm.
*/
void
xfs_ichgtime(
xfs_inode_t *ip,
int flags)
{
- struct inode *inode = LINVFS_GET_IP(XFS_ITOV(ip));
+ struct inode *inode = VFS_I(ip);
timespec_t tv;
+ int sync_it = 0;
- /*
- * We're not supposed to change timestamps in readonly-mounted
- * filesystems. Throw it away if anyone asks us.
- */
- if (unlikely(IS_RDONLY(inode)))
- return;
+ tv = current_fs_time(inode->i_sb);
- /*
- * Don't update access timestamps on reads if mounted "noatime".
- * Throw it away if anyone asks us.
- */
- if (unlikely(
- (ip->i_mount->m_flags & XFS_MOUNT_NOATIME || IS_NOATIME(inode)) &&
- (flags & (XFS_ICHGTIME_ACC|XFS_ICHGTIME_MOD|XFS_ICHGTIME_CHG)) ==
- XFS_ICHGTIME_ACC))
- return;
-
- nanotime(&tv);
- if (flags & XFS_ICHGTIME_MOD) {
+ if ((flags & XFS_ICHGTIME_MOD) &&
+ !timespec_equal(&inode->i_mtime, &tv)) {
inode->i_mtime = tv;
ip->i_d.di_mtime.t_sec = (__int32_t)tv.tv_sec;
ip->i_d.di_mtime.t_nsec = (__int32_t)tv.tv_nsec;
+ sync_it = 1;
}
- if (flags & XFS_ICHGTIME_ACC) {
- inode->i_atime = tv;
- ip->i_d.di_atime.t_sec = (__int32_t)tv.tv_sec;
- ip->i_d.di_atime.t_nsec = (__int32_t)tv.tv_nsec;
- }
- if (flags & XFS_ICHGTIME_CHG) {
+ if ((flags & XFS_ICHGTIME_CHG) &&
+ !timespec_equal(&inode->i_ctime, &tv)) {
inode->i_ctime = tv;
ip->i_d.di_ctime.t_sec = (__int32_t)tv.tv_sec;
ip->i_d.di_ctime.t_nsec = (__int32_t)tv.tv_nsec;
+ sync_it = 1;
}
/*
* ensure that the compiler does not reorder the update
* of i_update_core above the timestamp updates above.
*/
- SYNCHRONIZE();
- ip->i_update_core = 1;
- if (!(inode->i_state & I_LOCK))
+ if (sync_it) {
+ SYNCHRONIZE();
+ ip->i_update_core = 1;
mark_inode_dirty_sync(inode);
+ }
}
/*
- * Variant on the above which avoids querying the system clock
- * in situations where we know the Linux inode timestamps have
- * just been updated (and so we can update our inode cheaply).
- * We also skip the readonly and noatime checks here, they are
- * also catered for already.
+ * Hook in SELinux. This is not quite correct yet, what we really need
+ * here (as we do for default ACLs) is a mechanism by which creation of
+ * these attrs can be journalled at inode creation time (along with the
+ * inode, of course, such that log replay can't cause these to be lost).
*/
-void
-xfs_ichgtime_fast(
- xfs_inode_t *ip,
+STATIC int
+xfs_init_security(
struct inode *inode,
- int flags)
+ struct inode *dir)
{
- timespec_t *tvp;
-
- /*
- * We're not supposed to change timestamps in readonly-mounted
- * filesystems. Throw it away if anyone asks us.
- */
- if (unlikely(IS_RDONLY(inode)))
- return;
+ struct xfs_inode *ip = XFS_I(inode);
+ size_t length;
+ void *value;
+ char *name;
+ int error;
- /*
- * Don't update access timestamps on reads if mounted "noatime".
- * Throw it away if anyone asks us.
- */
- if (unlikely(
- (ip->i_mount->m_flags & XFS_MOUNT_NOATIME || IS_NOATIME(inode)) &&
- ((flags & (XFS_ICHGTIME_ACC|XFS_ICHGTIME_MOD|XFS_ICHGTIME_CHG)) ==
- XFS_ICHGTIME_ACC)))
- return;
-
- if (flags & XFS_ICHGTIME_MOD) {
- tvp = &inode->i_mtime;
- ip->i_d.di_mtime.t_sec = (__int32_t)tvp->tv_sec;
- ip->i_d.di_mtime.t_nsec = (__int32_t)tvp->tv_nsec;
- }
- if (flags & XFS_ICHGTIME_ACC) {
- tvp = &inode->i_atime;
- ip->i_d.di_atime.t_sec = (__int32_t)tvp->tv_sec;
- ip->i_d.di_atime.t_nsec = (__int32_t)tvp->tv_nsec;
- }
- if (flags & XFS_ICHGTIME_CHG) {
- tvp = &inode->i_ctime;
- ip->i_d.di_ctime.t_sec = (__int32_t)tvp->tv_sec;
- ip->i_d.di_ctime.t_nsec = (__int32_t)tvp->tv_nsec;
+ error = security_inode_init_security(inode, dir, &name,
+ &value, &length);
+ if (error) {
+ if (error == -EOPNOTSUPP)
+ return 0;
+ return -error;
}
- /*
- * We update the i_update_core field _after_ changing
- * the timestamps in order to coordinate properly with
- * xfs_iflush() so that we don't lose timestamp updates.
- * This keeps us from having to hold the inode lock
- * while doing this. We use the SYNCHRONIZE macro to
- * ensure that the compiler does not reorder the update
- * of i_update_core above the timestamp updates above.
- */
- SYNCHRONIZE();
- ip->i_update_core = 1;
- if (!(inode->i_state & I_LOCK))
- mark_inode_dirty_sync(inode);
+ error = xfs_attr_set(ip, name, value, length, ATTR_SECURE);
+ if (!error)
+ xfs_iflags_set(ip, XFS_IMODIFIED);
+
+ kfree(name);
+ kfree(value);
+ return error;
}
+static void
+xfs_dentry_to_name(
+ struct xfs_name *namep,
+ struct dentry *dentry)
+{
+ namep->name = dentry->d_name.name;
+ namep->len = dentry->d_name.len;
+}
-/*
- * Pull the link count and size up from the xfs inode to the linux inode
- */
STATIC void
-validate_fields(
- struct inode *ip)
+xfs_cleanup_inode(
+ struct inode *dir,
+ struct inode *inode,
+ struct dentry *dentry)
{
- vnode_t *vp = LINVFS_GET_VP(ip);
- vattr_t va;
- int error;
+ struct xfs_name teardown;
- va.va_mask = XFS_AT_NLINK|XFS_AT_SIZE|XFS_AT_NBLOCKS;
- VOP_GETATTR(vp, &va, ATTR_LAZY, NULL, error);
- if (likely(!error)) {
- ip->i_nlink = va.va_nlink;
- ip->i_blocks = va.va_nblocks;
-
- /* we're under i_mutex so i_size can't change under us */
- if (i_size_read(ip) != va.va_size)
- i_size_write(ip, va.va_size);
- }
-}
+ /* Oh, the horror.
+ * If we can't add the ACL or we fail in
+ * xfs_init_security we must back out.
+ * ENOSPC can hit here, among other things.
+ */
+ xfs_dentry_to_name(&teardown, dentry);
-/*
- * Determine whether a process has a valid fs_struct (kernel daemons
- * like knfsd don't have an fs_struct).
- *
- * XXX(hch): nfsd is broken, better fix it instead.
- */
-STATIC inline int
-has_fs_struct(struct task_struct *task)
-{
- return (task->fs != init_task.fs);
+ xfs_remove(XFS_I(dir), &teardown, XFS_I(inode));
+ iput(inode);
}
STATIC int
-linvfs_mknod(
+xfs_vn_mknod(
struct inode *dir,
struct dentry *dentry,
int mode,
dev_t rdev)
{
- struct inode *ip;
- vattr_t va;
- vnode_t *vp = NULL, *dvp = LINVFS_GET_VP(dir);
+ struct inode *inode;
+ struct xfs_inode *ip = NULL;
xfs_acl_t *default_acl = NULL;
- attrexists_t test_default_acl = _ACL_DEFAULT_EXISTS;
+ struct xfs_name name;
+ int (*test_default_acl)(struct inode *) = _ACL_DEFAULT_EXISTS;
int error;
/*
* Irix uses Missed'em'V split, but doesn't want to see
* the upper 5 bits of (14bit) major.
*/
- if (!sysv_valid_dev(rdev) || MAJOR(rdev) & ~0x1ff)
+ if (unlikely(!sysv_valid_dev(rdev) || MAJOR(rdev) & ~0x1ff))
return -EINVAL;
- if (test_default_acl && test_default_acl(dvp)) {
- if (!_ACL_ALLOC(default_acl))
+ if (test_default_acl && test_default_acl(dir)) {
+ if (!_ACL_ALLOC(default_acl)) {
return -ENOMEM;
- if (!_ACL_GET_DEFAULT(dvp, default_acl)) {
+ }
+ if (!_ACL_GET_DEFAULT(dir, default_acl)) {
_ACL_FREE(default_acl);
default_acl = NULL;
}
}
- if (IS_POSIXACL(dir) && !default_acl && has_fs_struct(current))
- mode &= ~current->fs->umask;
+ xfs_dentry_to_name(&name, dentry);
- memset(&va, 0, sizeof(va));
- va.va_mask = XFS_AT_TYPE|XFS_AT_MODE;
- va.va_mode = mode;
+ if (IS_POSIXACL(dir) && !default_acl)
+ mode &= ~current->fs->umask;
switch (mode & S_IFMT) {
- case S_IFCHR: case S_IFBLK: case S_IFIFO: case S_IFSOCK:
- va.va_rdev = sysv_encode_dev(rdev);
- va.va_mask |= XFS_AT_RDEV;
- /*FALLTHROUGH*/
+ case S_IFCHR:
+ case S_IFBLK:
+ case S_IFIFO:
+ case S_IFSOCK:
+ rdev = sysv_encode_dev(rdev);
case S_IFREG:
- VOP_CREATE(dvp, dentry, &va, &vp, NULL, error);
+ error = xfs_create(XFS_I(dir), &name, mode, rdev, &ip, NULL);
break;
case S_IFDIR:
- VOP_MKDIR(dvp, dentry, &va, &vp, NULL, error);
+ error = xfs_mkdir(XFS_I(dir), &name, mode, &ip, NULL);
break;
default:
error = EINVAL;
break;
}
+ if (unlikely(error))
+ goto out_free_acl;
+
+ inode = VFS_I(ip);
+
+ error = xfs_init_security(inode, dir);
+ if (unlikely(error))
+ goto out_cleanup_inode;
+
if (default_acl) {
- if (!error) {
- error = _ACL_INHERIT(vp, &va, default_acl);
- if (!error) {
- VMODIFY(vp);
- } else {
- struct dentry teardown = {};
- int err2;
-
- /* Oh, the horror.
- * If we can't add the ACL we must back out.
- * ENOSPC can hit here, among other things.
- */
- teardown.d_inode = ip = LINVFS_GET_IP(vp);
- teardown.d_name = dentry->d_name;
-
- vn_mark_bad(vp);
-
- if (S_ISDIR(mode))
- VOP_RMDIR(dvp, &teardown, NULL, err2);
- else
- VOP_REMOVE(dvp, &teardown, NULL, err2);
- VN_RELE(vp);
- }
- }
+ error = _ACL_INHERIT(inode, mode, default_acl);
+ if (unlikely(error))
+ goto out_cleanup_inode;
+ xfs_iflags_set(ip, XFS_IMODIFIED);
_ACL_FREE(default_acl);
}
- if (!error) {
- ASSERT(vp);
- ip = LINVFS_GET_IP(vp);
- if (S_ISCHR(mode) || S_ISBLK(mode))
- ip->i_rdev = rdev;
- else if (S_ISDIR(mode))
- validate_fields(ip);
- d_instantiate(dentry, ip);
- validate_fields(dir);
- }
+ d_instantiate(dentry, inode);
+ return -error;
+
+ out_cleanup_inode:
+ xfs_cleanup_inode(dir, inode, dentry);
+ out_free_acl:
+ if (default_acl)
+ _ACL_FREE(default_acl);
return -error;
}
STATIC int
-linvfs_create(
+xfs_vn_create(
struct inode *dir,
struct dentry *dentry,
int mode,
struct nameidata *nd)
{
- return linvfs_mknod(dir, dentry, mode, 0);
+ return xfs_vn_mknod(dir, dentry, mode, 0);
}
STATIC int
-linvfs_mkdir(
+xfs_vn_mkdir(
struct inode *dir,
struct dentry *dentry,
int mode)
{
- return linvfs_mknod(dir, dentry, mode|S_IFDIR, 0);
+ return xfs_vn_mknod(dir, dentry, mode|S_IFDIR, 0);
}
STATIC struct dentry *
-linvfs_lookup(
+xfs_vn_lookup(
struct inode *dir,
struct dentry *dentry,
struct nameidata *nd)
{
- struct vnode *vp = LINVFS_GET_VP(dir), *cvp;
+ struct xfs_inode *cip;
+ struct xfs_name name;
int error;
if (dentry->d_name.len >= MAXNAMELEN)
return ERR_PTR(-ENAMETOOLONG);
- VOP_LOOKUP(vp, dentry, &cvp, 0, NULL, NULL, error);
- if (error) {
+ xfs_dentry_to_name(&name, dentry);
+ error = xfs_lookup(XFS_I(dir), &name, &cip, NULL);
+ if (unlikely(error)) {
if (unlikely(error != ENOENT))
return ERR_PTR(-error);
d_add(dentry, NULL);
return NULL;
}
- return d_splice_alias(LINVFS_GET_IP(cvp), dentry);
+ return d_splice_alias(VFS_I(cip), dentry);
}
-STATIC int
-linvfs_link(
- struct dentry *old_dentry,
+STATIC struct dentry *
+xfs_vn_ci_lookup(
struct inode *dir,
- struct dentry *dentry)
+ struct dentry *dentry,
+ struct nameidata *nd)
{
- struct inode *ip; /* inode of guy being linked to */
- vnode_t *tdvp; /* target directory for new name/link */
- vnode_t *vp; /* vp of name being linked */
+ struct xfs_inode *ip;
+ struct xfs_name xname;
+ struct xfs_name ci_name;
+ struct qstr dname;
int error;
- ip = old_dentry->d_inode; /* inode being linked to */
- if (S_ISDIR(ip->i_mode))
- return -EPERM;
-
- tdvp = LINVFS_GET_VP(dir);
- vp = LINVFS_GET_VP(ip);
+ if (dentry->d_name.len >= MAXNAMELEN)
+ return ERR_PTR(-ENAMETOOLONG);
- VOP_LINK(tdvp, vp, dentry, NULL, error);
- if (!error) {
- VMODIFY(tdvp);
- VN_HOLD(vp);
- validate_fields(ip);
- d_instantiate(dentry, ip);
+ xfs_dentry_to_name(&xname, dentry);
+ error = xfs_lookup(XFS_I(dir), &xname, &ip, &ci_name);
+ if (unlikely(error)) {
+ if (unlikely(error != ENOENT))
+ return ERR_PTR(-error);
+ /*
+ * call d_add(dentry, NULL) here when d_drop_negative_children
+ * is called in xfs_vn_mknod (ie. allow negative dentries
+ * with CI filesystems).
+ */
+ return NULL;
}
- return -error;
+
+ /* if exact match, just splice and exit */
+ if (!ci_name.name)
+ return d_splice_alias(VFS_I(ip), dentry);
+
+ /* else case-insensitive match... */
+ dname.name = ci_name.name;
+ dname.len = ci_name.len;
+ dentry = d_add_ci(dentry, VFS_I(ip), &dname);
+ kmem_free(ci_name.name);
+ return dentry;
}
STATIC int
-linvfs_unlink(
+xfs_vn_link(
+ struct dentry *old_dentry,
struct inode *dir,
struct dentry *dentry)
{
- struct inode *inode;
- vnode_t *dvp; /* directory containing name to remove */
+ struct inode *inode; /* inode of guy being linked to */
+ struct xfs_name name;
int error;
- inode = dentry->d_inode;
- dvp = LINVFS_GET_VP(dir);
+ inode = old_dentry->d_inode;
+ xfs_dentry_to_name(&name, dentry);
- VOP_REMOVE(dvp, dentry, NULL, error);
- if (!error) {
- validate_fields(dir); /* For size only */
- validate_fields(inode);
+ igrab(inode);
+ error = xfs_link(XFS_I(dir), XFS_I(inode), &name);
+ if (unlikely(error)) {
+ iput(inode);
+ return -error;
}
- return -error;
+ xfs_iflags_set(XFS_I(dir), XFS_IMODIFIED);
+ d_instantiate(dentry, inode);
+ return 0;
}
STATIC int
-linvfs_symlink(
+xfs_vn_unlink(
struct inode *dir,
- struct dentry *dentry,
- const char *symname)
+ struct dentry *dentry)
{
- struct inode *ip;
- vattr_t va;
- vnode_t *dvp; /* directory containing name of symlink */
- vnode_t *cvp; /* used to lookup symlink to put in dentry */
+ struct xfs_name name;
int error;
- dvp = LINVFS_GET_VP(dir);
- cvp = NULL;
+ xfs_dentry_to_name(&name, dentry);
- memset(&va, 0, sizeof(va));
- va.va_mode = S_IFLNK |
- (irix_symlink_mode ? 0777 & ~current->fs->umask : S_IRWXUGO);
- va.va_mask = XFS_AT_TYPE|XFS_AT_MODE;
-
- error = 0;
- VOP_SYMLINK(dvp, dentry, &va, (char *)symname, &cvp, NULL, error);
- if (!error && cvp) {
- ip = LINVFS_GET_IP(cvp);
- d_instantiate(dentry, ip);
- validate_fields(dir);
- validate_fields(ip); /* size needs update */
- }
- return -error;
+ error = -xfs_remove(XFS_I(dir), &name, XFS_I(dentry->d_inode));
+ if (error)
+ return error;
+
+ /*
+ * With unlink, the VFS makes the dentry "negative": no inode,
+ * but still hashed. This is incompatible with case-insensitive
+ * mode, so invalidate (unhash) the dentry in CI-mode.
+ */
+ if (xfs_sb_version_hasasciici(&XFS_M(dir->i_sb)->m_sb))
+ d_invalidate(dentry);
+ return 0;
}
STATIC int
-linvfs_rmdir(
+xfs_vn_symlink(
struct inode *dir,
- struct dentry *dentry)
+ struct dentry *dentry,
+ const char *symname)
{
- struct inode *inode = dentry->d_inode;
- vnode_t *dvp = LINVFS_GET_VP(dir);
+ struct inode *inode;
+ struct xfs_inode *cip = NULL;
+ struct xfs_name name;
int error;
+ mode_t mode;
- VOP_RMDIR(dvp, dentry, NULL, error);
- if (!error) {
- validate_fields(inode);
- validate_fields(dir);
- }
+ mode = S_IFLNK |
+ (irix_symlink_mode ? 0777 & ~current->fs->umask : S_IRWXUGO);
+ xfs_dentry_to_name(&name, dentry);
+
+ error = xfs_symlink(XFS_I(dir), &name, symname, mode, &cip, NULL);
+ if (unlikely(error))
+ goto out;
+
+ inode = VFS_I(cip);
+
+ error = xfs_init_security(inode, dir);
+ if (unlikely(error))
+ goto out_cleanup_inode;
+
+ d_instantiate(dentry, inode);
+ return 0;
+
+ out_cleanup_inode:
+ xfs_cleanup_inode(dir, inode, dentry);
+ out:
return -error;
}
STATIC int
-linvfs_rename(
+xfs_vn_rename(
struct inode *odir,
struct dentry *odentry,
struct inode *ndir,
struct dentry *ndentry)
{
struct inode *new_inode = ndentry->d_inode;
- vnode_t *fvp; /* from directory */
- vnode_t *tvp; /* target directory */
- int error;
-
- fvp = LINVFS_GET_VP(odir);
- tvp = LINVFS_GET_VP(ndir);
+ struct xfs_name oname;
+ struct xfs_name nname;
- VOP_RENAME(fvp, odentry, tvp, ndentry, NULL, error);
- if (error)
- return -error;
+ xfs_dentry_to_name(&oname, odentry);
+ xfs_dentry_to_name(&nname, ndentry);
- if (new_inode)
- validate_fields(new_inode);
-
- validate_fields(odir);
- if (ndir != odir)
- validate_fields(ndir);
- return 0;
+ return -xfs_rename(XFS_I(odir), &oname, XFS_I(odentry->d_inode),
+ XFS_I(ndir), &nname, new_inode ?
+ XFS_I(new_inode) : NULL);
}
/*
* uio is kmalloced for this reason...
*/
STATIC void *
-linvfs_follow_link(
+xfs_vn_follow_link(
struct dentry *dentry,
struct nameidata *nd)
{
- vnode_t *vp;
- uio_t *uio;
- iovec_t iov;
- int error;
char *link;
+ int error = -ENOMEM;
- ASSERT(dentry);
- ASSERT(nd);
-
- link = (char *)kmalloc(MAXNAMELEN+1, GFP_KERNEL);
- if (!link) {
- nd_set_link(nd, ERR_PTR(-ENOMEM));
- return NULL;
- }
-
- uio = (uio_t *)kmalloc(sizeof(uio_t), GFP_KERNEL);
- if (!uio) {
- kfree(link);
- nd_set_link(nd, ERR_PTR(-ENOMEM));
- return NULL;
- }
-
- vp = LINVFS_GET_VP(dentry->d_inode);
-
- iov.iov_base = link;
- iov.iov_len = MAXNAMELEN;
+ link = kmalloc(MAXPATHLEN+1, GFP_KERNEL);
+ if (!link)
+ goto out_err;
- uio->uio_iov = &iov;
- uio->uio_offset = 0;
- uio->uio_segflg = UIO_SYSSPACE;
- uio->uio_resid = MAXNAMELEN;
- uio->uio_iovcnt = 1;
-
- VOP_READLINK(vp, uio, 0, NULL, error);
- if (error) {
- kfree(link);
- link = ERR_PTR(-error);
- } else {
- link[MAXNAMELEN - uio->uio_resid] = '\0';
- }
- kfree(uio);
+ error = -xfs_readlink(XFS_I(dentry->d_inode), link);
+ if (unlikely(error))
+ goto out_kfree;
nd_set_link(nd, link);
return NULL;
+
+ out_kfree:
+ kfree(link);
+ out_err:
+ nd_set_link(nd, ERR_PTR(error));
+ return NULL;
}
STATIC void
-linvfs_put_link(
+xfs_vn_put_link(
struct dentry *dentry,
struct nameidata *nd,
void *p)
#ifdef CONFIG_XFS_POSIX_ACL
STATIC int
-linvfs_permission(
- struct inode *inode,
- int mode,
- struct nameidata *nd)
+xfs_check_acl(
+ struct inode *inode,
+ int mask)
{
- vnode_t *vp = LINVFS_GET_VP(inode);
- int error;
+ struct xfs_inode *ip = XFS_I(inode);
+ int error;
- mode <<= 6; /* convert from linux to vnode access bits */
- VOP_ACCESS(vp, mode, NULL, error);
- return -error;
+ xfs_itrace_entry(ip);
+
+ if (XFS_IFORK_Q(ip)) {
+ error = xfs_acl_iaccess(ip, mask, NULL);
+ if (error != -1)
+ return -error;
+ }
+
+ return -EAGAIN;
+}
+
+STATIC int
+xfs_vn_permission(
+ struct inode *inode,
+ int mask)
+{
+ return generic_permission(inode, mask, xfs_check_acl);
}
#else
-#define linvfs_permission NULL
+#define xfs_vn_permission NULL
#endif
STATIC int
-linvfs_getattr(
- struct vfsmount *mnt,
- struct dentry *dentry,
- struct kstat *stat)
+xfs_vn_getattr(
+ struct vfsmount *mnt,
+ struct dentry *dentry,
+ struct kstat *stat)
{
- struct inode *inode = dentry->d_inode;
- vnode_t *vp = LINVFS_GET_VP(inode);
- int error = 0;
+ struct inode *inode = dentry->d_inode;
+ struct xfs_inode *ip = XFS_I(inode);
+ struct xfs_mount *mp = ip->i_mount;
+
+ xfs_itrace_entry(ip);
+
+ if (XFS_FORCED_SHUTDOWN(mp))
+ return XFS_ERROR(EIO);
+
+ stat->size = XFS_ISIZE(ip);
+ stat->dev = inode->i_sb->s_dev;
+ stat->mode = ip->i_d.di_mode;
+ stat->nlink = ip->i_d.di_nlink;
+ stat->uid = ip->i_d.di_uid;
+ stat->gid = ip->i_d.di_gid;
+ stat->ino = ip->i_ino;
+#if XFS_BIG_INUMS
+ stat->ino += mp->m_inoadd;
+#endif
+ stat->atime = inode->i_atime;
+ stat->mtime.tv_sec = ip->i_d.di_mtime.t_sec;
+ stat->mtime.tv_nsec = ip->i_d.di_mtime.t_nsec;
+ stat->ctime.tv_sec = ip->i_d.di_ctime.t_sec;
+ stat->ctime.tv_nsec = ip->i_d.di_ctime.t_nsec;
+ stat->blocks =
+ XFS_FSB_TO_BB(mp, ip->i_d.di_nblocks + ip->i_delayed_blks);
+
+
+ switch (inode->i_mode & S_IFMT) {
+ case S_IFBLK:
+ case S_IFCHR:
+ stat->blksize = BLKDEV_IOSIZE;
+ stat->rdev = MKDEV(sysv_major(ip->i_df.if_u2.if_rdev) & 0x1ff,
+ sysv_minor(ip->i_df.if_u2.if_rdev));
+ break;
+ default:
+ if (XFS_IS_REALTIME_INODE(ip)) {
+ /*
+ * If the file blocks are being allocated from a
+ * realtime volume, then return the inode's realtime
+ * extent size or the realtime volume's extent size.
+ */
+ stat->blksize =
+ xfs_get_extsz_hint(ip) << mp->m_sb.sb_blocklog;
+ } else
+ stat->blksize = xfs_preferred_iosize(mp);
+ stat->rdev = 0;
+ break;
+ }
- if (unlikely(vp->v_flag & VMODIFIED))
- error = vn_revalidate(vp);
- if (!error)
- generic_fillattr(inode, stat);
return 0;
}
STATIC int
-linvfs_setattr(
+xfs_vn_setattr(
struct dentry *dentry,
- struct iattr *attr)
+ struct iattr *iattr)
{
- struct inode *inode = dentry->d_inode;
- unsigned int ia_valid = attr->ia_valid;
- vnode_t *vp = LINVFS_GET_VP(inode);
- vattr_t vattr;
- int flags = 0;
- int error;
-
- memset(&vattr, 0, sizeof(vattr_t));
- if (ia_valid & ATTR_UID) {
- vattr.va_mask |= XFS_AT_UID;
- vattr.va_uid = attr->ia_uid;
- }
- if (ia_valid & ATTR_GID) {
- vattr.va_mask |= XFS_AT_GID;
- vattr.va_gid = attr->ia_gid;
- }
- if (ia_valid & ATTR_SIZE) {
- vattr.va_mask |= XFS_AT_SIZE;
- vattr.va_size = attr->ia_size;
- }
- if (ia_valid & ATTR_ATIME) {
- vattr.va_mask |= XFS_AT_ATIME;
- vattr.va_atime = attr->ia_atime;
- }
- if (ia_valid & ATTR_MTIME) {
- vattr.va_mask |= XFS_AT_MTIME;
- vattr.va_mtime = attr->ia_mtime;
- }
- if (ia_valid & ATTR_CTIME) {
- vattr.va_mask |= XFS_AT_CTIME;
- vattr.va_ctime = attr->ia_ctime;
- }
- if (ia_valid & ATTR_MODE) {
- vattr.va_mask |= XFS_AT_MODE;
- vattr.va_mode = attr->ia_mode;
- if (!in_group_p(inode->i_gid) && !capable(CAP_FSETID))
- inode->i_mode &= ~S_ISGID;
- }
-
- if (ia_valid & (ATTR_MTIME_SET | ATTR_ATIME_SET))
- flags |= ATTR_UTIME;
-#ifdef ATTR_NO_BLOCK
- if ((ia_valid & ATTR_NO_BLOCK))
- flags |= ATTR_NONBLOCK;
-#endif
-
- VOP_SETATTR(vp, &vattr, flags, NULL, error);
- if (error)
- return -error;
- vn_revalidate(vp);
- return error;
+ return -xfs_setattr(XFS_I(dentry->d_inode), iattr, 0, NULL);
}
+/*
+ * block_truncate_page can return an error, but we can't propagate it
+ * at all here. Leave a complaint + stack trace in the syslog because
+ * this could be bad. If it is bad, we need to propagate the error further.
+ */
STATIC void
-linvfs_truncate(
+xfs_vn_truncate(
struct inode *inode)
{
- block_truncate_page(inode->i_mapping, inode->i_size, linvfs_get_block);
+ int error;
+ error = block_truncate_page(inode->i_mapping, inode->i_size,
+ xfs_get_blocks);
+ WARN_ON(error);
}
-STATIC int
-linvfs_setxattr(
- struct dentry *dentry,
- const char *name,
- const void *data,
- size_t size,
- int flags)
+STATIC long
+xfs_vn_fallocate(
+ struct inode *inode,
+ int mode,
+ loff_t offset,
+ loff_t len)
{
- vnode_t *vp = LINVFS_GET_VP(dentry->d_inode);
- char *attr = (char *)name;
- attrnames_t *namesp;
- int xflags = 0;
- int error;
-
- namesp = attr_lookup_namespace(attr, attr_namespaces, ATTR_NAMECOUNT);
- if (!namesp)
- return -EOPNOTSUPP;
- attr += namesp->attr_namelen;
- error = namesp->attr_capable(vp, NULL);
- if (error)
- return error;
+ long error;
+ loff_t new_size = 0;
+ xfs_flock64_t bf;
+ xfs_inode_t *ip = XFS_I(inode);
+
+ /* preallocation on directories not yet supported */
+ error = -ENODEV;
+ if (S_ISDIR(inode->i_mode))
+ goto out_error;
+
+ bf.l_whence = 0;
+ bf.l_start = offset;
+ bf.l_len = len;
+
+ xfs_ilock(ip, XFS_IOLOCK_EXCL);
+ error = xfs_change_file_space(ip, XFS_IOC_RESVSP, &bf,
+ 0, NULL, XFS_ATTR_NOLOCK);
+ if (!error && !(mode & FALLOC_FL_KEEP_SIZE) &&
+ offset + len > i_size_read(inode))
+ new_size = offset + len;
+
+ /* Change file size if needed */
+ if (new_size) {
+ struct iattr iattr;
+
+ iattr.ia_valid = ATTR_SIZE;
+ iattr.ia_size = new_size;
+ error = xfs_setattr(ip, &iattr, XFS_ATTR_NOLOCK, NULL);
+ }
- /* Convert Linux syscall to XFS internal ATTR flags */
- if (flags & XATTR_CREATE)
- xflags |= ATTR_CREATE;
- if (flags & XATTR_REPLACE)
- xflags |= ATTR_REPLACE;
- xflags |= namesp->attr_flag;
- return namesp->attr_set(vp, attr, (void *)data, size, xflags);
+ xfs_iunlock(ip, XFS_IOLOCK_EXCL);
+out_error:
+ return error;
}
-STATIC ssize_t
-linvfs_getxattr(
- struct dentry *dentry,
- const char *name,
- void *data,
- size_t size)
-{
- vnode_t *vp = LINVFS_GET_VP(dentry->d_inode);
- char *attr = (char *)name;
- attrnames_t *namesp;
- int xflags = 0;
- ssize_t error;
-
- namesp = attr_lookup_namespace(attr, attr_namespaces, ATTR_NAMECOUNT);
- if (!namesp)
- return -EOPNOTSUPP;
- attr += namesp->attr_namelen;
- error = namesp->attr_capable(vp, NULL);
- if (error)
- return error;
-
- /* Convert Linux syscall to XFS internal ATTR flags */
- if (!size) {
- xflags |= ATTR_KERNOVAL;
- data = NULL;
- }
- xflags |= namesp->attr_flag;
- return namesp->attr_get(vp, attr, (void *)data, size, xflags);
-}
+static const struct inode_operations xfs_inode_operations = {
+ .permission = xfs_vn_permission,
+ .truncate = xfs_vn_truncate,
+ .getattr = xfs_vn_getattr,
+ .setattr = xfs_vn_setattr,
+ .setxattr = generic_setxattr,
+ .getxattr = generic_getxattr,
+ .removexattr = generic_removexattr,
+ .listxattr = xfs_vn_listxattr,
+ .fallocate = xfs_vn_fallocate,
+};
-STATIC ssize_t
-linvfs_listxattr(
- struct dentry *dentry,
- char *data,
- size_t size)
-{
- vnode_t *vp = LINVFS_GET_VP(dentry->d_inode);
- int error, xflags = ATTR_KERNAMELS;
- ssize_t result;
+static const struct inode_operations xfs_dir_inode_operations = {
+ .create = xfs_vn_create,
+ .lookup = xfs_vn_lookup,
+ .link = xfs_vn_link,
+ .unlink = xfs_vn_unlink,
+ .symlink = xfs_vn_symlink,
+ .mkdir = xfs_vn_mkdir,
+ /*
+ * Yes, XFS uses the same method for rmdir and unlink.
+ *
+ * There are some subtile differences deeper in the code,
+ * but we use S_ISDIR to check for those.
+ */
+ .rmdir = xfs_vn_unlink,
+ .mknod = xfs_vn_mknod,
+ .rename = xfs_vn_rename,
+ .permission = xfs_vn_permission,
+ .getattr = xfs_vn_getattr,
+ .setattr = xfs_vn_setattr,
+ .setxattr = generic_setxattr,
+ .getxattr = generic_getxattr,
+ .removexattr = generic_removexattr,
+ .listxattr = xfs_vn_listxattr,
+};
- if (!size)
- xflags |= ATTR_KERNOVAL;
- xflags |= capable(CAP_SYS_ADMIN) ? ATTR_KERNFULLS : ATTR_KERNORMALS;
+static const struct inode_operations xfs_dir_ci_inode_operations = {
+ .create = xfs_vn_create,
+ .lookup = xfs_vn_ci_lookup,
+ .link = xfs_vn_link,
+ .unlink = xfs_vn_unlink,
+ .symlink = xfs_vn_symlink,
+ .mkdir = xfs_vn_mkdir,
+ /*
+ * Yes, XFS uses the same method for rmdir and unlink.
+ *
+ * There are some subtile differences deeper in the code,
+ * but we use S_ISDIR to check for those.
+ */
+ .rmdir = xfs_vn_unlink,
+ .mknod = xfs_vn_mknod,
+ .rename = xfs_vn_rename,
+ .permission = xfs_vn_permission,
+ .getattr = xfs_vn_getattr,
+ .setattr = xfs_vn_setattr,
+ .setxattr = generic_setxattr,
+ .getxattr = generic_getxattr,
+ .removexattr = generic_removexattr,
+ .listxattr = xfs_vn_listxattr,
+};
- error = attr_generic_list(vp, data, size, xflags, &result);
- if (error < 0)
- return error;
- return result;
-}
+static const struct inode_operations xfs_symlink_inode_operations = {
+ .readlink = generic_readlink,
+ .follow_link = xfs_vn_follow_link,
+ .put_link = xfs_vn_put_link,
+ .permission = xfs_vn_permission,
+ .getattr = xfs_vn_getattr,
+ .setattr = xfs_vn_setattr,
+ .setxattr = generic_setxattr,
+ .getxattr = generic_getxattr,
+ .removexattr = generic_removexattr,
+ .listxattr = xfs_vn_listxattr,
+};
-STATIC int
-linvfs_removexattr(
- struct dentry *dentry,
- const char *name)
+STATIC void
+xfs_diflags_to_iflags(
+ struct inode *inode,
+ struct xfs_inode *ip)
{
- vnode_t *vp = LINVFS_GET_VP(dentry->d_inode);
- char *attr = (char *)name;
- attrnames_t *namesp;
- int xflags = 0;
- int error;
-
- namesp = attr_lookup_namespace(attr, attr_namespaces, ATTR_NAMECOUNT);
- if (!namesp)
- return -EOPNOTSUPP;
- attr += namesp->attr_namelen;
- error = namesp->attr_capable(vp, NULL);
- if (error)
- return error;
- xflags |= namesp->attr_flag;
- return namesp->attr_remove(vp, attr, xflags);
+ if (ip->i_d.di_flags & XFS_DIFLAG_IMMUTABLE)
+ inode->i_flags |= S_IMMUTABLE;
+ else
+ inode->i_flags &= ~S_IMMUTABLE;
+ if (ip->i_d.di_flags & XFS_DIFLAG_APPEND)
+ inode->i_flags |= S_APPEND;
+ else
+ inode->i_flags &= ~S_APPEND;
+ if (ip->i_d.di_flags & XFS_DIFLAG_SYNC)
+ inode->i_flags |= S_SYNC;
+ else
+ inode->i_flags &= ~S_SYNC;
+ if (ip->i_d.di_flags & XFS_DIFLAG_NOATIME)
+ inode->i_flags |= S_NOATIME;
+ else
+ inode->i_flags &= ~S_NOATIME;
}
+/*
+ * Initialize the Linux inode, set up the operation vectors and
+ * unlock the inode.
+ *
+ * When reading existing inodes from disk this is called directly
+ * from xfs_iget, when creating a new inode it is called from
+ * xfs_ialloc after setting up the inode.
+ */
+void
+xfs_setup_inode(
+ struct xfs_inode *ip)
+{
+ struct inode *inode = ip->i_vnode;
+
+ inode->i_mode = ip->i_d.di_mode;
+ inode->i_nlink = ip->i_d.di_nlink;
+ inode->i_uid = ip->i_d.di_uid;
+ inode->i_gid = ip->i_d.di_gid;
+
+ switch (inode->i_mode & S_IFMT) {
+ case S_IFBLK:
+ case S_IFCHR:
+ inode->i_rdev =
+ MKDEV(sysv_major(ip->i_df.if_u2.if_rdev) & 0x1ff,
+ sysv_minor(ip->i_df.if_u2.if_rdev));
+ break;
+ default:
+ inode->i_rdev = 0;
+ break;
+ }
-struct inode_operations linvfs_file_inode_operations = {
- .permission = linvfs_permission,
- .truncate = linvfs_truncate,
- .getattr = linvfs_getattr,
- .setattr = linvfs_setattr,
- .setxattr = linvfs_setxattr,
- .getxattr = linvfs_getxattr,
- .listxattr = linvfs_listxattr,
- .removexattr = linvfs_removexattr,
-};
+ inode->i_generation = ip->i_d.di_gen;
+ i_size_write(inode, ip->i_d.di_size);
+ inode->i_atime.tv_sec = ip->i_d.di_atime.t_sec;
+ inode->i_atime.tv_nsec = ip->i_d.di_atime.t_nsec;
+ inode->i_mtime.tv_sec = ip->i_d.di_mtime.t_sec;
+ inode->i_mtime.tv_nsec = ip->i_d.di_mtime.t_nsec;
+ inode->i_ctime.tv_sec = ip->i_d.di_ctime.t_sec;
+ inode->i_ctime.tv_nsec = ip->i_d.di_ctime.t_nsec;
+ xfs_diflags_to_iflags(inode, ip);
+ xfs_iflags_clear(ip, XFS_IMODIFIED);
+
+ switch (inode->i_mode & S_IFMT) {
+ case S_IFREG:
+ inode->i_op = &xfs_inode_operations;
+ inode->i_fop = &xfs_file_operations;
+ inode->i_mapping->a_ops = &xfs_address_space_operations;
+ break;
+ case S_IFDIR:
+ if (xfs_sb_version_hasasciici(&XFS_M(inode->i_sb)->m_sb))
+ inode->i_op = &xfs_dir_ci_inode_operations;
+ else
+ inode->i_op = &xfs_dir_inode_operations;
+ inode->i_fop = &xfs_dir_file_operations;
+ break;
+ case S_IFLNK:
+ inode->i_op = &xfs_symlink_inode_operations;
+ if (!(ip->i_df.if_flags & XFS_IFINLINE))
+ inode->i_mapping->a_ops = &xfs_address_space_operations;
+ break;
+ default:
+ inode->i_op = &xfs_inode_operations;
+ init_special_inode(inode, inode->i_mode, inode->i_rdev);
+ break;
+ }
-struct inode_operations linvfs_dir_inode_operations = {
- .create = linvfs_create,
- .lookup = linvfs_lookup,
- .link = linvfs_link,
- .unlink = linvfs_unlink,
- .symlink = linvfs_symlink,
- .mkdir = linvfs_mkdir,
- .rmdir = linvfs_rmdir,
- .mknod = linvfs_mknod,
- .rename = linvfs_rename,
- .permission = linvfs_permission,
- .getattr = linvfs_getattr,
- .setattr = linvfs_setattr,
- .setxattr = linvfs_setxattr,
- .getxattr = linvfs_getxattr,
- .listxattr = linvfs_listxattr,
- .removexattr = linvfs_removexattr,
-};
+ xfs_iflags_clear(ip, XFS_INEW);
+ barrier();
-struct inode_operations linvfs_symlink_inode_operations = {
- .readlink = generic_readlink,
- .follow_link = linvfs_follow_link,
- .put_link = linvfs_put_link,
- .permission = linvfs_permission,
- .getattr = linvfs_getattr,
- .setattr = linvfs_setattr,
- .setxattr = linvfs_setxattr,
- .getxattr = linvfs_getxattr,
- .listxattr = linvfs_listxattr,
- .removexattr = linvfs_removexattr,
-};
+ unlock_new_inode(inode);
+}
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff