diff options
Diffstat (limited to 'fs/xfs/xfs_iops.c')
-rw-r--r-- | fs/xfs/xfs_iops.c | 291 |
1 files changed, 168 insertions, 123 deletions
diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c index 27e0e544e963..72129493e9d3 100644 --- a/fs/xfs/xfs_iops.c +++ b/fs/xfs/xfs_iops.c @@ -39,6 +39,7 @@ #include "xfs_da_btree.h" #include "xfs_dir2_priv.h" #include "xfs_dinode.h" +#include "xfs_trans_space.h" #include <linux/capability.h> #include <linux/xattr.h> @@ -48,6 +49,18 @@ #include <linux/fiemap.h> #include <linux/slab.h> +/* + * Directories have different lock order w.r.t. mmap_sem compared to regular + * files. This is due to readdir potentially triggering page faults on a user + * buffer inside filldir(), and this happens with the ilock on the directory + * held. For regular files, the lock order is the other way around - the + * mmap_sem is taken during the page fault, and then we lock the ilock to do + * block mapping. Hence we need a different class for the directory ilock so + * that lockdep can tell them apart. + */ +static struct lock_class_key xfs_nondir_ilock_class; +static struct lock_class_key xfs_dir_ilock_class; + static int xfs_initxattrs( struct inode *inode, @@ -60,7 +73,7 @@ xfs_initxattrs( for (xattr = xattr_array; xattr->name != NULL; xattr++) { error = xfs_attr_set(ip, xattr->name, xattr->value, - xattr->value_len, ATTR_SECURE); + xattr->value_len, ATTR_SECURE); if (error < 0) break; } @@ -81,7 +94,7 @@ xfs_init_security( const struct qstr *qstr) { return security_inode_init_security(inode, dir, qstr, - &xfs_initxattrs, NULL); + &xfs_initxattrs, NULL); } static void @@ -111,19 +124,19 @@ xfs_cleanup_inode( xfs_dentry_to_name(&teardown, dentry, 0); xfs_remove(XFS_I(dir), &teardown, XFS_I(inode)); - iput(inode); } STATIC int -xfs_vn_mknod( +xfs_generic_create( struct inode *dir, struct dentry *dentry, umode_t mode, - dev_t rdev) + dev_t rdev, + bool tmpfile) /* unnamed file */ { struct inode *inode; struct xfs_inode *ip = NULL; - struct posix_acl *default_acl = NULL; + struct posix_acl *default_acl, *acl; struct xfs_name name; int error; @@ -139,17 +152,16 @@ xfs_vn_mknod( rdev = 0; } - if (IS_POSIXACL(dir)) { - default_acl = xfs_get_acl(dir, ACL_TYPE_DEFAULT); - if (IS_ERR(default_acl)) - return PTR_ERR(default_acl); + error = posix_acl_create(dir, &mode, &default_acl, &acl); + if (error) + return error; - if (!default_acl) - mode &= ~current_umask(); + if (!tmpfile) { + xfs_dentry_to_name(&name, dentry, mode); + error = xfs_create(XFS_I(dir), &name, mode, rdev, &ip); + } else { + error = xfs_create_tmpfile(XFS_I(dir), dentry, mode, &ip); } - - xfs_dentry_to_name(&name, dentry, mode); - error = xfs_create(XFS_I(dir), &name, mode, rdev, &ip); if (unlikely(error)) goto out_free_acl; @@ -159,22 +171,46 @@ xfs_vn_mknod( if (unlikely(error)) goto out_cleanup_inode; +#ifdef CONFIG_XFS_POSIX_ACL if (default_acl) { - error = -xfs_inherit_acl(inode, default_acl); - default_acl = NULL; - if (unlikely(error)) + error = xfs_set_acl(inode, default_acl, ACL_TYPE_DEFAULT); + if (error) + goto out_cleanup_inode; + } + if (acl) { + error = xfs_set_acl(inode, acl, ACL_TYPE_ACCESS); + if (error) goto out_cleanup_inode; } +#endif + if (tmpfile) + d_tmpfile(dentry, inode); + else + d_instantiate(dentry, inode); - d_instantiate(dentry, inode); - return -error; + out_free_acl: + if (default_acl) + posix_acl_release(default_acl); + if (acl) + posix_acl_release(acl); + return error; out_cleanup_inode: - xfs_cleanup_inode(dir, inode, dentry); - out_free_acl: - posix_acl_release(default_acl); - return -error; + if (!tmpfile) + xfs_cleanup_inode(dir, inode, dentry); + iput(inode); + goto out_free_acl; +} + +STATIC int +xfs_vn_mknod( + struct inode *dir, + struct dentry *dentry, + umode_t mode, + dev_t rdev) +{ + return xfs_generic_create(dir, dentry, mode, rdev, false); } STATIC int @@ -212,8 +248,8 @@ xfs_vn_lookup( xfs_dentry_to_name(&name, dentry, 0); error = xfs_lookup(XFS_I(dir), &name, &cip, NULL); if (unlikely(error)) { - if (unlikely(error != ENOENT)) - return ERR_PTR(-error); + if (unlikely(error != -ENOENT)) + return ERR_PTR(error); d_add(dentry, NULL); return NULL; } @@ -239,8 +275,8 @@ xfs_vn_ci_lookup( xfs_dentry_to_name(&xname, dentry, 0); error = xfs_lookup(XFS_I(dir), &xname, &ip, &ci_name); if (unlikely(error)) { - if (unlikely(error != ENOENT)) - return ERR_PTR(-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 @@ -275,7 +311,7 @@ xfs_vn_link( error = xfs_link(XFS_I(dir), XFS_I(inode), &name); if (unlikely(error)) - return -error; + return error; ihold(inode); d_instantiate(dentry, inode); @@ -292,7 +328,7 @@ xfs_vn_unlink( xfs_dentry_to_name(&name, dentry, 0); - error = -xfs_remove(XFS_I(dir), &name, XFS_I(dentry->d_inode)); + error = xfs_remove(XFS_I(dir), &name, XFS_I(dentry->d_inode)); if (error) return error; @@ -337,8 +373,9 @@ xfs_vn_symlink( out_cleanup_inode: xfs_cleanup_inode(dir, inode, dentry); + iput(inode); out: - return -error; + return error; } STATIC int @@ -355,8 +392,8 @@ xfs_vn_rename( xfs_dentry_to_name(&oname, odentry, 0); xfs_dentry_to_name(&nname, ndentry, odentry->d_inode->i_mode); - return -xfs_rename(XFS_I(odir), &oname, XFS_I(odentry->d_inode), - XFS_I(ndir), &nname, new_inode ? + return xfs_rename(XFS_I(odir), &oname, XFS_I(odentry->d_inode), + XFS_I(ndir), &nname, new_inode ? XFS_I(new_inode) : NULL); } @@ -377,7 +414,7 @@ xfs_vn_follow_link( if (!link) goto out_err; - error = -xfs_readlink(XFS_I(dentry->d_inode), link); + error = xfs_readlink(XFS_I(dentry->d_inode), link); if (unlikely(error)) goto out_kfree; @@ -391,18 +428,6 @@ xfs_vn_follow_link( return NULL; } -STATIC void -xfs_vn_put_link( - struct dentry *dentry, - struct nameidata *nd, - void *p) -{ - char *s = nd_get_link(nd); - - if (!IS_ERR(s)) - kfree(s); -} - STATIC int xfs_vn_getattr( struct vfsmount *mnt, @@ -416,7 +441,7 @@ xfs_vn_getattr( trace_xfs_getattr(ip); if (XFS_FORCED_SHUTDOWN(mp)) - return -XFS_ERROR(EIO); + return -EIO; stat->size = XFS_ISIZE(ip); stat->dev = inode->i_sb->s_dev; @@ -459,14 +484,12 @@ xfs_vn_getattr( static void xfs_setattr_mode( - struct xfs_trans *tp, struct xfs_inode *ip, struct iattr *iattr) { - struct inode *inode = VFS_I(ip); - umode_t mode = iattr->ia_mode; + struct inode *inode = VFS_I(ip); + umode_t mode = iattr->ia_mode; - ASSERT(tp); ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); ip->i_d.di_mode &= S_IFMT; @@ -476,6 +499,32 @@ xfs_setattr_mode( inode->i_mode |= mode & ~S_IFMT; } +static void +xfs_setattr_time( + struct xfs_inode *ip, + struct iattr *iattr) +{ + struct inode *inode = VFS_I(ip); + + ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); + + if (iattr->ia_valid & ATTR_ATIME) { + inode->i_atime = iattr->ia_atime; + ip->i_d.di_atime.t_sec = iattr->ia_atime.tv_sec; + ip->i_d.di_atime.t_nsec = iattr->ia_atime.tv_nsec; + } + if (iattr->ia_valid & ATTR_CTIME) { + inode->i_ctime = iattr->ia_ctime; + ip->i_d.di_ctime.t_sec = iattr->ia_ctime.tv_sec; + ip->i_d.di_ctime.t_nsec = iattr->ia_ctime.tv_nsec; + } + if (iattr->ia_valid & ATTR_MTIME) { + inode->i_mtime = iattr->ia_mtime; + ip->i_d.di_mtime.t_sec = iattr->ia_mtime.tv_sec; + ip->i_d.di_mtime.t_nsec = iattr->ia_mtime.tv_nsec; + } +} + int xfs_setattr_nonsize( struct xfs_inode *ip, @@ -497,14 +546,14 @@ xfs_setattr_nonsize( /* If acls are being inherited, we already have this checked */ if (!(flags & XFS_ATTR_NOACL)) { if (mp->m_flags & XFS_MOUNT_RDONLY) - return XFS_ERROR(EROFS); + return -EROFS; if (XFS_FORCED_SHUTDOWN(mp)) - return XFS_ERROR(EIO); + return -EIO; - error = -inode_change_ok(inode, iattr); + error = inode_change_ok(inode, iattr); if (error) - return XFS_ERROR(error); + return error; } ASSERT((mask & ATTR_SIZE) == 0); @@ -618,7 +667,8 @@ xfs_setattr_nonsize( } if (!gid_eq(igid, gid)) { if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_GQUOTA_ON(mp)) { - ASSERT(!XFS_IS_PQUOTA_ON(mp)); + ASSERT(xfs_sb_version_has_pquotino(&mp->m_sb) || + !XFS_IS_PQUOTA_ON(mp)); ASSERT(mask & ATTR_GID); ASSERT(gdqp); olddquot2 = xfs_qm_vop_chown(tp, ip, @@ -629,30 +679,10 @@ xfs_setattr_nonsize( } } - /* - * Change file access modes. - */ if (mask & ATTR_MODE) - xfs_setattr_mode(tp, ip, iattr); - - /* - * Change file access or modified times. - */ - if (mask & ATTR_ATIME) { - inode->i_atime = iattr->ia_atime; - ip->i_d.di_atime.t_sec = iattr->ia_atime.tv_sec; - ip->i_d.di_atime.t_nsec = iattr->ia_atime.tv_nsec; - } - if (mask & ATTR_CTIME) { - inode->i_ctime = iattr->ia_ctime; - ip->i_d.di_ctime.t_sec = iattr->ia_ctime.tv_sec; - ip->i_d.di_ctime.t_nsec = iattr->ia_ctime.tv_nsec; - } - if (mask & ATTR_MTIME) { - inode->i_mtime = iattr->ia_mtime; - ip->i_d.di_mtime.t_sec = iattr->ia_mtime.tv_sec; - ip->i_d.di_mtime.t_nsec = iattr->ia_mtime.tv_nsec; - } + xfs_setattr_mode(ip, iattr); + if (mask & (ATTR_ATIME|ATTR_CTIME|ATTR_MTIME)) + xfs_setattr_time(ip, iattr); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); @@ -673,7 +703,7 @@ xfs_setattr_nonsize( xfs_qm_dqrele(gdqp); if (error) - return XFS_ERROR(error); + return error; /* * XXX(hch): Updating the ACL entries is not atomic vs the i_mode @@ -683,9 +713,9 @@ xfs_setattr_nonsize( * Posix ACL code seems to care about this issue either. */ if ((mask & ATTR_MODE) && !(flags & XFS_ATTR_NOACL)) { - error = -xfs_acl_chmod(inode); + error = posix_acl_chmod(inode, inode->i_mode); if (error) - return XFS_ERROR(error); + return error; } return 0; @@ -709,7 +739,6 @@ xfs_setattr_size( { struct xfs_mount *mp = ip->i_mount; struct inode *inode = VFS_I(ip); - int mask = iattr->ia_valid; xfs_off_t oldsize, newsize; struct xfs_trans *tp; int error; @@ -719,19 +748,19 @@ xfs_setattr_size( trace_xfs_setattr(ip); if (mp->m_flags & XFS_MOUNT_RDONLY) - return XFS_ERROR(EROFS); + return -EROFS; if (XFS_FORCED_SHUTDOWN(mp)) - return XFS_ERROR(EIO); + return -EIO; - error = -inode_change_ok(inode, iattr); + error = inode_change_ok(inode, iattr); if (error) - return XFS_ERROR(error); + return error; ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL)); ASSERT(S_ISREG(ip->i_d.di_mode)); - ASSERT((mask & (ATTR_UID|ATTR_GID|ATTR_ATIME|ATTR_ATIME_SET| - ATTR_MTIME_SET|ATTR_KILL_PRIV|ATTR_TIMES_SET)) == 0); + ASSERT((iattr->ia_valid & (ATTR_UID|ATTR_GID|ATTR_ATIME|ATTR_ATIME_SET| + ATTR_MTIME_SET|ATTR_KILL_PRIV|ATTR_TIMES_SET)) == 0); oldsize = inode->i_size; newsize = iattr->ia_size; @@ -740,7 +769,7 @@ xfs_setattr_size( * Short circuit the truncate case for zero length files. */ if (newsize == 0 && oldsize == 0 && ip->i_d.di_nextents == 0) { - if (!(mask & (ATTR_CTIME|ATTR_MTIME))) + if (!(iattr->ia_valid & (ATTR_CTIME|ATTR_MTIME))) return 0; /* @@ -789,7 +818,7 @@ xfs_setattr_size( * care about here. */ if (oldsize != ip->i_d.di_size && newsize > ip->i_d.di_size) { - error = -filemap_write_and_wait_range(VFS_I(ip)->i_mapping, + error = filemap_write_and_wait_range(VFS_I(ip)->i_mapping, ip->i_d.di_size, newsize); if (error) return error; @@ -800,22 +829,34 @@ xfs_setattr_size( */ inode_dio_wait(inode); - error = -block_truncate_page(inode->i_mapping, newsize, xfs_get_blocks); + /* + * Do all the page cache truncate work outside the transaction context + * as the "lock" order is page lock->log space reservation. i.e. + * locking pages inside the transaction can ABBA deadlock with + * writeback. We have to do the VFS inode size update before we truncate + * the pagecache, however, to avoid racing with page faults beyond the + * new EOF they are not serialised against truncate operations except by + * page locks and size updates. + * + * Hence we are in a situation where a truncate can fail with ENOMEM + * from xfs_trans_reserve(), but having already truncated the in-memory + * version of the file (i.e. made user visible changes). There's not + * much we can do about this, except to hope that the caller sees ENOMEM + * and retries the truncate operation. + */ + error = block_truncate_page(inode->i_mapping, newsize, xfs_get_blocks); if (error) return error; + truncate_setsize(inode, newsize); tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_SIZE); error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0); if (error) goto out_trans_cancel; - truncate_setsize(inode, newsize); - commit_flags = XFS_TRANS_RELEASE_LOG_RES; lock_flags |= XFS_ILOCK_EXCL; - xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_trans_ijoin(tp, ip, 0); /* @@ -828,10 +869,11 @@ xfs_setattr_size( * these flags set. For all other operations the VFS set these flags * explicitly if it wants a timestamp update. */ - if (newsize != oldsize && (!(mask & (ATTR_CTIME | ATTR_MTIME)))) { + if (newsize != oldsize && + !(iattr->ia_valid & (ATTR_CTIME | ATTR_MTIME))) { iattr->ia_ctime = iattr->ia_mtime = current_fs_time(inode->i_sb); - mask |= ATTR_CTIME | ATTR_MTIME; + iattr->ia_valid |= ATTR_CTIME | ATTR_MTIME; } /* @@ -867,22 +909,10 @@ xfs_setattr_size( xfs_inode_clear_eofblocks_tag(ip); } - /* - * Change file access modes. - */ - if (mask & ATTR_MODE) - xfs_setattr_mode(tp, ip, iattr); - - if (mask & ATTR_CTIME) { - inode->i_ctime = iattr->ia_ctime; - ip->i_d.di_ctime.t_sec = iattr->ia_ctime.tv_sec; - ip->i_d.di_ctime.t_nsec = iattr->ia_ctime.tv_nsec; - } - if (mask & ATTR_MTIME) { - inode->i_mtime = iattr->ia_mtime; - ip->i_d.di_mtime.t_sec = iattr->ia_mtime.tv_sec; - ip->i_d.di_mtime.t_nsec = iattr->ia_mtime.tv_nsec; - } + if (iattr->ia_valid & ATTR_MODE) + xfs_setattr_mode(ip, iattr); + if (iattr->ia_valid & (ATTR_ATIME|ATTR_CTIME|ATTR_MTIME)) + xfs_setattr_time(ip, iattr); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); @@ -920,7 +950,7 @@ xfs_vn_setattr( error = xfs_setattr_nonsize(ip, iattr, 0); } - return -error; + return error; } STATIC int @@ -940,7 +970,7 @@ xfs_vn_update_time( error = xfs_trans_reserve(tp, &M_RES(mp)->tr_fsyncts, 0, 0); if (error) { xfs_trans_cancel(tp, 0); - return -error; + return error; } xfs_ilock(ip, XFS_ILOCK_EXCL); @@ -961,7 +991,7 @@ xfs_vn_update_time( } xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); xfs_trans_log_inode(tp, ip, XFS_ILOG_TIMESTAMP); - return -xfs_trans_commit(tp, 0); + return xfs_trans_commit(tp, 0); } #define XFS_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR) @@ -1006,7 +1036,7 @@ xfs_fiemap_format( *full = 1; /* user array now full */ } - return -error; + return error; } STATIC int @@ -1025,12 +1055,12 @@ xfs_vn_fiemap( return error; /* Set up bmap header for xfs internal routine */ - bm.bmv_offset = BTOBB(start); + bm.bmv_offset = BTOBBT(start); /* Special case for whole file */ if (length == FIEMAP_MAX_OFFSET) bm.bmv_length = -1LL; else - bm.bmv_length = BTOBB(length); + bm.bmv_length = BTOBB(start + length) - bm.bmv_offset; /* We add one because in getbmap world count includes the header */ bm.bmv_count = !fieinfo->fi_extents_max ? MAXEXTNUM : @@ -1045,13 +1075,23 @@ xfs_vn_fiemap( error = xfs_getbmap(ip, &bm, xfs_fiemap_format, fieinfo); if (error) - return -error; + return error; return 0; } +STATIC int +xfs_vn_tmpfile( + struct inode *dir, + struct dentry *dentry, + umode_t mode) +{ + return xfs_generic_create(dir, dentry, mode, 0, true); +} + static const struct inode_operations xfs_inode_operations = { .get_acl = xfs_get_acl, + .set_acl = xfs_set_acl, .getattr = xfs_vn_getattr, .setattr = xfs_vn_setattr, .setxattr = generic_setxattr, @@ -1079,6 +1119,7 @@ static const struct inode_operations xfs_dir_inode_operations = { .mknod = xfs_vn_mknod, .rename = xfs_vn_rename, .get_acl = xfs_get_acl, + .set_acl = xfs_set_acl, .getattr = xfs_vn_getattr, .setattr = xfs_vn_setattr, .setxattr = generic_setxattr, @@ -1086,6 +1127,7 @@ static const struct inode_operations xfs_dir_inode_operations = { .removexattr = generic_removexattr, .listxattr = xfs_vn_listxattr, .update_time = xfs_vn_update_time, + .tmpfile = xfs_vn_tmpfile, }; static const struct inode_operations xfs_dir_ci_inode_operations = { @@ -1105,6 +1147,7 @@ static const struct inode_operations xfs_dir_ci_inode_operations = { .mknod = xfs_vn_mknod, .rename = xfs_vn_rename, .get_acl = xfs_get_acl, + .set_acl = xfs_set_acl, .getattr = xfs_vn_getattr, .setattr = xfs_vn_setattr, .setxattr = generic_setxattr, @@ -1112,13 +1155,13 @@ static const struct inode_operations xfs_dir_ci_inode_operations = { .removexattr = generic_removexattr, .listxattr = xfs_vn_listxattr, .update_time = xfs_vn_update_time, + .tmpfile = xfs_vn_tmpfile, }; static const struct inode_operations xfs_symlink_inode_operations = { .readlink = generic_readlink, .follow_link = xfs_vn_follow_link, - .put_link = xfs_vn_put_link, - .get_acl = xfs_get_acl, + .put_link = kfree_put_link, .getattr = xfs_vn_getattr, .setattr = xfs_vn_setattr, .setxattr = generic_setxattr, @@ -1205,6 +1248,7 @@ xfs_setup_inode( xfs_diflags_to_iflags(inode, ip); ip->d_ops = ip->i_mount->m_nondir_inode_ops; + lockdep_set_class(&ip->i_lock.mr_lock, &xfs_nondir_ilock_class); switch (inode->i_mode & S_IFMT) { case S_IFREG: inode->i_op = &xfs_inode_operations; @@ -1212,6 +1256,7 @@ xfs_setup_inode( inode->i_mapping->a_ops = &xfs_address_space_operations; break; case S_IFDIR: + lockdep_set_class(&ip->i_lock.mr_lock, &xfs_dir_ilock_class); if (xfs_sb_version_hasasciici(&XFS_M(inode->i_sb)->m_sb)) inode->i_op = &xfs_dir_ci_inode_operations; else |