Re: [PATCH 15/32] IGET: Stop EXT4 from using iget() and read_inode() [try #2]

[Jan Kara <jack@suse.cz>]

> Stop the EXT4 filesystem from using iget() and read_inode().  Replace
> ext4_read_inode() with ext4_iget(), and call that instead of iget().
> ext4_iget() then uses iget_locked() directly and returns a proper error code
> instead of an inode in the event of an error.
> 
> ext4_fill_super() returns any error incurred when getting the root inode
> instead of EINVAL.
> 
> Signed-off-by: David Howells <dhowells@redhat.com>
  Acked-by: Jan Kara <jack@suse.cz>

> ---
> 
>  fs/ext4/ialloc.c        |   58 ++++++++++++++++++++++++++++-------------------
>  fs/ext4/inode.c         |   25 +++++++++++++++-----
>  fs/ext4/namei.c         |   29 +++++++-----------------
>  fs/ext4/resize.c        |    7 ++----
>  fs/ext4/super.c         |   36 ++++++++++++++++-------------
>  include/linux/ext4_fs.h |    2 +-
>  6 files changed, 87 insertions(+), 70 deletions(-)
> 
> diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
> index 427f830..4fba007 100644
> --- a/fs/ext4/ialloc.c
> +++ b/fs/ext4/ialloc.c
> @@ -659,14 +659,15 @@ struct inode *ext4_orphan_get(struct super_block *sb, unsigned long ino)
>  	unsigned long max_ino = le32_to_cpu(EXT4_SB(sb)->s_es->s_inodes_count);
>  	unsigned long block_group;
>  	int bit;
> -	struct buffer_head *bitmap_bh = NULL;
> +	struct buffer_head *bitmap_bh;
>  	struct inode *inode = NULL;
> +	long err = -EIO;
>  
>  	/* Error cases - e2fsck has already cleaned up for us */
>  	if (ino > max_ino) {
>  		ext4_warning(sb, __FUNCTION__,
>  			     "bad orphan ino %lu!  e2fsck was run?", ino);
> -		goto out;
> +		goto error;
>  	}
>  
>  	block_group = (ino - 1) / EXT4_INODES_PER_GROUP(sb);
> @@ -675,38 +676,49 @@ struct inode *ext4_orphan_get(struct super_block *sb, unsigned long ino)
>  	if (!bitmap_bh) {
>  		ext4_warning(sb, __FUNCTION__,
>  			     "inode bitmap error for orphan %lu", ino);
> -		goto out;
> +		goto error;
>  	}
>  
>  	/* Having the inode bit set should be a 100% indicator that this
>  	 * is a valid orphan (no e2fsck run on fs).  Orphans also include
>  	 * inodes that were being truncated, so we can't check i_nlink==0.
>  	 */
> -	if (!ext4_test_bit(bit, bitmap_bh->b_data) ||
> -			!(inode = iget(sb, ino)) || is_bad_inode(inode) ||
> -			NEXT_ORPHAN(inode) > max_ino) {
> -		ext4_warning(sb, __FUNCTION__,
> -			     "bad orphan inode %lu!  e2fsck was run?", ino);
> -		printk(KERN_NOTICE "ext4_test_bit(bit=%d, block=%llu) = %d\n",
> -		       bit, (unsigned long long)bitmap_bh->b_blocknr,
> -		       ext4_test_bit(bit, bitmap_bh->b_data));
> -		printk(KERN_NOTICE "inode=%p\n", inode);
> -		if (inode) {
> -			printk(KERN_NOTICE "is_bad_inode(inode)=%d\n",
> -			       is_bad_inode(inode));
> -			printk(KERN_NOTICE "NEXT_ORPHAN(inode)=%u\n",
> -			       NEXT_ORPHAN(inode));
> -			printk(KERN_NOTICE "max_ino=%lu\n", max_ino);
> -		}
> +	if (!ext4_test_bit(bit, bitmap_bh->b_data))
> +		goto bad_orphan;
> +
> +	inode = ext4_iget(sb, ino);
> +	if (IS_ERR(inode))
> +		goto iget_failed;
> +
> +	if (NEXT_ORPHAN(inode) > max_ino)
> +		goto bad_orphan;
> +	brelse(bitmap_bh);
> +	return inode;
> +
> +iget_failed:
> +	err = PTR_ERR(inode);
> +	inode = NULL;
> +bad_orphan:
> +	ext4_warning(sb, __FUNCTION__,
> +		     "bad orphan inode %lu!  e2fsck was run?", ino);
> +	printk(KERN_NOTICE "ext4_test_bit(bit=%d, block=%llu) = %d\n",
> +	       bit, (unsigned long long)bitmap_bh->b_blocknr,
> +	       ext4_test_bit(bit, bitmap_bh->b_data));
> +	printk(KERN_NOTICE "inode=%p\n", inode);
> +	if (inode) {
> +		printk(KERN_NOTICE "is_bad_inode(inode)=%d\n",
> +		       is_bad_inode(inode));
> +		printk(KERN_NOTICE "NEXT_ORPHAN(inode)=%u\n",
> +		       NEXT_ORPHAN(inode));
> +		printk(KERN_NOTICE "max_ino=%lu\n", max_ino);
>  		/* Avoid freeing blocks if we got a bad deleted inode */
> -		if (inode && inode->i_nlink == 0)
> +		if (inode->i_nlink == 0)
>  			inode->i_blocks = 0;
>  		iput(inode);
> -		inode = NULL;
>  	}
> -out:
>  	brelse(bitmap_bh);
> -	return inode;
> +error:
> +	return ERR_PTR(err);
>  }
>  
>  unsigned long ext4_count_free_inodes (struct super_block * sb)
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index a4848e0..c4fb1eb 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -2595,21 +2595,31 @@ void ext4_get_inode_flags(struct ext4_inode_info *ei)
>  		ei->i_flags |= EXT4_DIRSYNC_FL;
>  }
>  
> -void ext4_read_inode(struct inode * inode)
> +struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
>  {
>  	struct ext4_iloc iloc;
>  	struct ext4_inode *raw_inode;
> -	struct ext4_inode_info *ei = EXT4_I(inode);
> +	struct ext4_inode_info *ei;
>  	struct buffer_head *bh;
> +	struct inode *inode;
> +	long ret;
>  	int block;
>  
> +	inode = iget_locked(sb, ino);
> +	if (!inode)
> +		return ERR_PTR(-ENOMEM);
> +	if (!(inode->i_state & I_NEW))
> +		return inode;
> +
> +	ei = EXT4_I(inode);
>  #ifdef CONFIG_EXT4DEV_FS_POSIX_ACL
>  	ei->i_acl = EXT4_ACL_NOT_CACHED;
>  	ei->i_default_acl = EXT4_ACL_NOT_CACHED;
>  #endif
>  	ei->i_block_alloc_info = NULL;
>  
> -	if (__ext4_get_inode_loc(inode, &iloc, 0))
> +	ret = __ext4_get_inode_loc(inode, &iloc, 0);
> +	if (ret < 0)
>  		goto bad_inode;
>  	bh = iloc.bh;
>  	raw_inode = ext4_raw_inode(&iloc);
> @@ -2636,6 +2646,7 @@ void ext4_read_inode(struct inode * inode)
>  		    !(EXT4_SB(inode->i_sb)->s_mount_state & EXT4_ORPHAN_FS)) {
>  			/* this inode is deleted */
>  			brelse (bh);
> +			ret = -ESTALE;
>  			goto bad_inode;
>  		}
>  		/* The only unlinked inodes we let through here have
> @@ -2683,6 +2694,7 @@ void ext4_read_inode(struct inode * inode)
>  		if (EXT4_GOOD_OLD_INODE_SIZE + ei->i_extra_isize >
>  		    EXT4_INODE_SIZE(inode->i_sb)) {
>  			brelse (bh);
> +			ret = -EIO;
>  			goto bad_inode;
>  		}
>  		if (ei->i_extra_isize == 0) {
> @@ -2729,11 +2741,12 @@ void ext4_read_inode(struct inode * inode)
>  	}
>  	brelse (iloc.bh);
>  	ext4_set_inode_flags(inode);
> -	return;
> +	unlock_new_inode(inode);
> +	return inode;
>  
>  bad_inode:
> -	make_bad_inode(inode);
> -	return;
> +	iget_failed(inode);
> +	return ERR_PTR(ret);
>  }
>  
>  /*
> diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
> index 5fdb862..301f41f 100644
> --- a/fs/ext4/namei.c
> +++ b/fs/ext4/namei.c
> @@ -1044,17 +1044,11 @@ static struct dentry *ext4_lookup(struct inode * dir, struct dentry *dentry, str
>  		if (!ext4_valid_inum(dir->i_sb, ino)) {
>  			ext4_error(dir->i_sb, "ext4_lookup",
>  				   "bad inode number: %lu", ino);
> -			inode = NULL;
> -		} else
> -			inode = iget(dir->i_sb, ino);
> -
> -		if (!inode)
> -			return ERR_PTR(-EACCES);
> -
> -		if (is_bad_inode(inode)) {
> -			iput(inode);
> -			return ERR_PTR(-ENOENT);
> +			return ERR_PTR(-EIO);
>  		}
> +		inode = ext4_iget(dir->i_sb, ino);
> +		if (IS_ERR(inode))
> +			return ERR_CAST(inode);
>  	}
>  	return d_splice_alias(inode, dentry);
>  }
> @@ -1083,18 +1077,13 @@ struct dentry *ext4_get_parent(struct dentry *child)
>  	if (!ext4_valid_inum(child->d_inode->i_sb, ino)) {
>  		ext4_error(child->d_inode->i_sb, "ext4_get_parent",
>  			   "bad inode number: %lu", ino);
> -		inode = NULL;
> -	} else
> -		inode = iget(child->d_inode->i_sb, ino);
> -
> -	if (!inode)
> -		return ERR_PTR(-EACCES);
> -
> -	if (is_bad_inode(inode)) {
> -		iput(inode);
> -		return ERR_PTR(-ENOENT);
> +		return ERR_PTR(-EIO);
>  	}
>  
> +	inode = ext4_iget(child->d_inode->i_sb, ino);
> +	if (IS_ERR(inode))
> +		return ERR_CAST(inode);
> +
>  	parent = d_alloc_anon(inode);
>  	if (!parent) {
>  		iput(inode);
> diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c
> index aa11d7d..345f901 100644
> --- a/fs/ext4/resize.c
> +++ b/fs/ext4/resize.c
> @@ -757,12 +757,11 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input)
>  				     "No reserved GDT blocks, can't resize");
>  			return -EPERM;
>  		}
> -		inode = iget(sb, EXT4_RESIZE_INO);
> -		if (!inode || is_bad_inode(inode)) {
> +		inode = ext4_iget(sb, EXT4_RESIZE_INO);
> +		if (IS_ERR(inode)) {
>  			ext4_warning(sb, __FUNCTION__,
>  				     "Error opening resize inode");
> -			iput(inode);
> -			return -ENOENT;
> +			return PTR_ERR(inode);
>  		}
>  	}
>  
> diff --git a/fs/ext4/super.c b/fs/ext4/super.c
> index 3c1397f..05434a1 100644
> --- a/fs/ext4/super.c
> +++ b/fs/ext4/super.c
> @@ -634,11 +634,10 @@ static struct dentry *ext4_get_dentry(struct super_block *sb, void *vobjp)
>  	 * Currently we don't know the generation for parent directory, so
>  	 * a generation of 0 means "accept any"
>  	 */
> -	inode = iget(sb, ino);
> -	if (inode == NULL)
> -		return ERR_PTR(-ENOMEM);
> -	if (is_bad_inode(inode) ||
> -	    (generation && inode->i_generation != generation)) {
> +	inode = ext4_iget(sb, ino);
> +	if (IS_ERR(inode))
> +		return ERR_CAST(inode);
> +	if (generation && inode->i_generation != generation) {
>  		iput(inode);
>  		return ERR_PTR(-ESTALE);
>  	}
> @@ -700,7 +699,6 @@ static struct quotactl_ops ext4_qctl_operations = {
>  static const struct super_operations ext4_sops = {
>  	.alloc_inode	= ext4_alloc_inode,
>  	.destroy_inode	= ext4_destroy_inode,
> -	.read_inode	= ext4_read_inode,
>  	.write_inode	= ext4_write_inode,
>  	.dirty_inode	= ext4_dirty_inode,
>  	.delete_inode	= ext4_delete_inode,
> @@ -1472,6 +1470,7 @@ static int ext4_fill_super (struct super_block *sb, void *data, int silent)
>  	unsigned long journal_devnum = 0;
>  	unsigned long def_mount_opts;
>  	struct inode *root;
> +	int ret = -EINVAL;
>  	int blocksize;
>  	int hblock;
>  	int db_count;
> @@ -1862,19 +1861,24 @@ static int ext4_fill_super (struct super_block *sb, void *data, int silent)
>  	 * so we can safely mount the rest of the filesystem now.
>  	 */
>  
> -	root = iget(sb, EXT4_ROOT_INO);
> -	sb->s_root = d_alloc_root(root);
> -	if (!sb->s_root) {
> +	root = ext4_iget(sb, EXT4_ROOT_INO);
> +	if (IS_ERR(root)) {
>  		printk(KERN_ERR "EXT4-fs: get root inode failed\n");
> -		iput(root);
> +		ret = PTR_ERR(root);
>  		goto failed_mount4;
>  	}
>  	if (!S_ISDIR(root->i_mode) || !root->i_blocks || !root->i_size) {
> -		dput(sb->s_root);
> -		sb->s_root = NULL;
> +		iput(root);
>  		printk(KERN_ERR "EXT4-fs: corrupt root inode, run e2fsck\n");
>  		goto failed_mount4;
>  	}
> +	sb->s_root = d_alloc_root(root);
> +	if (!sb->s_root) {
> +		printk(KERN_ERR "EXT4-fs: get root dentry failed\n");
> +		iput(root);
> +		ret = -ENOMEM;
> +		goto failed_mount4;
> +	}
>  
>  	ext4_setup_super (sb, es, sb->s_flags & MS_RDONLY);
>  
> @@ -1954,7 +1958,7 @@ out_fail:
>  	sb->s_fs_info = NULL;
>  	kfree(sbi);
>  	lock_kernel();
> -	return -EINVAL;
> +	return ret;
>  }
>  
>  /*
> @@ -1990,8 +1994,8 @@ static journal_t *ext4_get_journal(struct super_block *sb,
>  	 * things happen if we iget() an unused inode, as the subsequent
>  	 * iput() will try to delete it. */
>  
> -	journal_inode = iget(sb, journal_inum);
> -	if (!journal_inode) {
> +	journal_inode = ext4_iget(sb, journal_inum);
> +	if (IS_ERR(journal_inode)) {
>  		printk(KERN_ERR "EXT4-fs: no journal found.\n");
>  		return NULL;
>  	}
> @@ -2004,7 +2008,7 @@ static journal_t *ext4_get_journal(struct super_block *sb,
>  
>  	jbd_debug(2, "Journal inode found at %p: %Ld bytes\n",
>  		  journal_inode, journal_inode->i_size);
> -	if (is_bad_inode(journal_inode) || !S_ISREG(journal_inode->i_mode)) {
> +	if (!S_ISREG(journal_inode->i_mode)) {
>  		printk(KERN_ERR "EXT4-fs: invalid journal inode.\n");
>  		iput(journal_inode);
>  		return NULL;
> diff --git a/include/linux/ext4_fs.h b/include/linux/ext4_fs.h
> index cdee7aa..12354d5 100644
> --- a/include/linux/ext4_fs.h
> +++ b/include/linux/ext4_fs.h
> @@ -944,7 +944,7 @@ int ext4_get_blocks_handle(handle_t *handle, struct inode *inode,
>  	sector_t iblock, unsigned long maxblocks, struct buffer_head *bh_result,
>  	int create, int extend_disksize);
>  
> -extern void ext4_read_inode (struct inode *);
> +extern struct inode *ext4_iget(struct super_block *, unsigned long);
>  extern int  ext4_write_inode (struct inode *, int);
>  extern int  ext4_setattr (struct dentry *, struct iattr *);
>  extern void ext4_delete_inode (struct inode *);
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
-- 
Jan Kara <jack@suse.cz>
SuSE CR Labs