1 files changed, 81 insertions, 21 deletions

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index a084e84610f8..7335cff90a7d 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -196,7 +196,12 @@ void ext4_evict_inode(struct inode *inode)
 {
 	handle_t *handle;
 	int err;
-	int extra_credits = 3;
+	/*
+	 * Credits for final inode cleanup and freeing:
+	 * sb + inode (ext4_orphan_del()), block bitmap, group descriptor
+	 * (xattr block freeing), bitmap, group descriptor (inode freeing)
+	 */
+	int extra_credits = 6;
 	struct ext4_xattr_inode_array *ea_inode_array = NULL;
 
 	trace_ext4_evict_inode(inode);
@@ -252,8 +257,12 @@ void ext4_evict_inode(struct inode *inode)
 	if (!IS_NOQUOTA(inode))
 		extra_credits += EXT4_MAXQUOTAS_DEL_BLOCKS(inode->i_sb);
 
+	/*
+	 * Block bitmap, group descriptor, and inode are accounted in both
+	 * ext4_blocks_for_truncate() and extra_credits. So subtract 3.
+	 */
 	handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE,
-				 ext4_blocks_for_truncate(inode)+extra_credits);
+			 ext4_blocks_for_truncate(inode) + extra_credits - 3);
 	if (IS_ERR(handle)) {
 		ext4_std_error(inode->i_sb, PTR_ERR(handle));
 		/*
@@ -2145,7 +2154,7 @@ static int ext4_writepage(struct page *page,
 	bool keep_towrite = false;
 
 	if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) {
-		ext4_invalidatepage(page, 0, PAGE_SIZE);
+		inode->i_mapping->a_ops->invalidatepage(page, 0, PAGE_SIZE);
 		unlock_page(page);
 		return -EIO;
 	}
@@ -2582,7 +2591,7 @@ update_disksize:
 	 * truncate are avoided by checking i_size under i_data_sem.
 	 */
 	disksize = ((loff_t)mpd->first_page) << PAGE_SHIFT;
-	if (disksize > EXT4_I(inode)->i_disksize) {
+	if (disksize > READ_ONCE(EXT4_I(inode)->i_disksize)) {
 		int err2;
 		loff_t i_size;
 
@@ -2743,7 +2752,7 @@ static int ext4_writepages(struct address_space *mapping,
 	if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb))))
 		return -EIO;
 
-	percpu_down_read(&sbi->s_journal_flag_rwsem);
+	percpu_down_read(&sbi->s_writepages_rwsem);
 	trace_ext4_writepages(inode, wbc);
 
 	/*
@@ -2964,7 +2973,7 @@ unplug:
 out_writepages:
 	trace_ext4_writepages_result(inode, wbc, ret,
 				     nr_to_write - wbc->nr_to_write);
-	percpu_up_read(&sbi->s_journal_flag_rwsem);
+	percpu_up_read(&sbi->s_writepages_rwsem);
 	return ret;
 }
 
@@ -2979,13 +2988,13 @@ static int ext4_dax_writepages(struct address_space *mapping,
 	if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb))))
 		return -EIO;
 
-	percpu_down_read(&sbi->s_journal_flag_rwsem);
+	percpu_down_read(&sbi->s_writepages_rwsem);
 	trace_ext4_writepages(inode, wbc);
 
 	ret = dax_writeback_mapping_range(mapping, inode->i_sb->s_bdev, wbc);
 	trace_ext4_writepages_result(inode, wbc, ret,
 				     nr_to_write - wbc->nr_to_write);
-	percpu_up_read(&sbi->s_journal_flag_rwsem);
+	percpu_up_read(&sbi->s_writepages_rwsem);
 	return ret;
 }
 
@@ -3559,8 +3568,14 @@ retry:
 			return ret;
 	}
 
+	/*
+	 * Writes that span EOF might trigger an I/O size update on completion,
+	 * so consider them to be dirty for the purposes of O_DSYNC, even if
+	 * there is no other metadata changes being made or are pending here.
+	 */
 	iomap->flags = 0;
-	if (ext4_inode_datasync_dirty(inode))
+	if (ext4_inode_datasync_dirty(inode) ||
+	    offset + length > i_size_read(inode))
 		iomap->flags |= IOMAP_F_DIRTY;
 	iomap->bdev = inode->i_sb->s_bdev;
 	iomap->dax_dev = sbi->s_daxdev;
@@ -3857,13 +3872,24 @@ static ssize_t ext4_direct_IO_read(struct kiocb *iocb, struct iov_iter *iter)
 	struct inode *inode = mapping->host;
 	size_t count = iov_iter_count(iter);
 	ssize_t ret;
+	loff_t offset = iocb->ki_pos;
+	loff_t size = i_size_read(inode);
+
+	if (offset >= size)
+		return 0;
 
 	/*
 	 * Shared inode_lock is enough for us - it protects against concurrent
 	 * writes & truncates and since we take care of writing back page cache,
 	 * we are protected against page writeback as well.
 	 */
-	inode_lock_shared(inode);
+	if (iocb->ki_flags & IOCB_NOWAIT) {
+		if (!inode_trylock_shared(inode))
+			return -EAGAIN;
+	} else {
+		inode_lock_shared(inode);
+	}
+
 	ret = filemap_write_and_wait_range(mapping, iocb->ki_pos,
 					   iocb->ki_pos + count - 1);
 	if (ret)
@@ -4692,7 +4718,7 @@ make_io:
 			if (end > table)
 				end = table;
 			while (b <= end)
-				sb_breadahead(sb, b++);
+				sb_breadahead_unmovable(sb, b++);
 		}
 
 		/*
@@ -4979,6 +5005,18 @@ struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
 		ret = -EFSCORRUPTED;
 		goto bad_inode;
 	}
+	/*
+	 * If dir_index is not enabled but there's dir with INDEX flag set,
+	 * we'd normally treat htree data as empty space. But with metadata
+	 * checksumming that corrupts checksums so forbid that.
+	 */
+	if (!ext4_has_feature_dir_index(sb) && ext4_has_metadata_csum(sb) &&
+	    ext4_test_inode_flag(inode, EXT4_INODE_INDEX)) {
+		ext4_error_inode(inode, function, line, 0,
+			 "iget: Dir with htree data on filesystem without dir_index feature.");
+		ret = -EFSCORRUPTED;
+		goto bad_inode;
+	}
 	ei->i_disksize = inode->i_size;
 #ifdef CONFIG_QUOTA
 	ei->i_reserved_quota = 0;
@@ -5116,6 +5154,9 @@ struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
 				 "iget: bogus i_mode (%o)", inode->i_mode);
 		goto bad_inode;
 	}
+	if (IS_CASEFOLDED(inode) && !ext4_has_feature_casefold(inode->i_sb))
+		ext4_error_inode(inode, function, line, 0,
+				 "casefold flag without casefold feature");
 	brelse(iloc.bh);
 
 	unlock_new_inode(inode);
@@ -5132,7 +5173,7 @@ static int ext4_inode_blocks_set(handle_t *handle,
 				struct ext4_inode_info *ei)
 {
 	struct inode *inode = &(ei->vfs_inode);
-	u64 i_blocks = inode->i_blocks;
+	u64 i_blocks = READ_ONCE(inode->i_blocks);
 	struct super_block *sb = inode->i_sb;
 
 	if (i_blocks <= ~0U) {
@@ -5475,11 +5516,15 @@ static void ext4_wait_for_tail_page_commit(struct inode *inode)
 
 	offset = inode->i_size & (PAGE_SIZE - 1);
 	/*
-	 * All buffers in the last page remain valid? Then there's nothing to
-	 * do. We do the check mainly to optimize the common PAGE_SIZE ==
-	 * blocksize case
+	 * If the page is fully truncated, we don't need to wait for any commit
+	 * (and we even should not as __ext4_journalled_invalidatepage() may
+	 * strip all buffers from the page but keep the page dirty which can then
+	 * confuse e.g. concurrent ext4_writepage() seeing dirty page without
+	 * buffers). Also we don't need to wait for any commit if all buffers in
+	 * the page remain valid. This is most beneficial for the common case of
+	 * blocksize == PAGESIZE.
 	 */
-	if (offset > PAGE_SIZE - i_blocksize(inode))
+	if (!offset || offset > (PAGE_SIZE - i_blocksize(inode)))
 		return;
 	while (1) {
 		page = find_lock_page(inode->i_mapping,
@@ -5929,8 +5974,23 @@ static int __ext4_expand_extra_isize(struct inode *inode,
 {
 	struct ext4_inode *raw_inode;
 	struct ext4_xattr_ibody_header *header;
+	unsigned int inode_size = EXT4_INODE_SIZE(inode->i_sb);
+	struct ext4_inode_info *ei = EXT4_I(inode);
 	int error;
 
+	/* this was checked at iget time, but double check for good measure */
+	if ((EXT4_GOOD_OLD_INODE_SIZE + ei->i_extra_isize > inode_size) ||
+	    (ei->i_extra_isize & 3)) {
+		EXT4_ERROR_INODE(inode, "bad extra_isize %u (inode size %u)",
+				 ei->i_extra_isize,
+				 EXT4_INODE_SIZE(inode->i_sb));
+		return -EFSCORRUPTED;
+	}
+	if ((new_extra_isize < ei->i_extra_isize) ||
+	    (new_extra_isize < 4) ||
+	    (new_extra_isize > inode_size - EXT4_GOOD_OLD_INODE_SIZE))
+		return -EINVAL;	/* Should never happen */
+
 	raw_inode = ext4_raw_inode(iloc);
 
 	header = IHDR(inode, raw_inode);
@@ -6024,7 +6084,7 @@ int ext4_expand_extra_isize(struct inode *inode,
 	error = ext4_journal_get_write_access(handle, iloc->bh);
 	if (error) {
 		brelse(iloc->bh);
-		goto out_stop;
+		goto out_unlock;
 	}
 
 	error = __ext4_expand_extra_isize(inode, new_extra_isize, iloc,
@@ -6034,8 +6094,8 @@ int ext4_expand_extra_isize(struct inode *inode,
 	if (!error)
 		error = rc;
 
+out_unlock:
 	ext4_write_unlock_xattr(inode, &no_expand);
-out_stop:
 	ext4_journal_stop(handle);
 	return error;
 }
@@ -6150,7 +6210,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val)
 		}
 	}
 
-	percpu_down_write(&sbi->s_journal_flag_rwsem);
+	percpu_down_write(&sbi->s_writepages_rwsem);
 	jbd2_journal_lock_updates(journal);
 
 	/*
@@ -6167,7 +6227,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val)
 		err = jbd2_journal_flush(journal);
 		if (err < 0) {
 			jbd2_journal_unlock_updates(journal);
-			percpu_up_write(&sbi->s_journal_flag_rwsem);
+			percpu_up_write(&sbi->s_writepages_rwsem);
 			return err;
 		}
 		ext4_clear_inode_flag(inode, EXT4_INODE_JOURNAL_DATA);
@@ -6175,7 +6235,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val)
 	ext4_set_aops(inode);
 
 	jbd2_journal_unlock_updates(journal);
-	percpu_up_write(&sbi->s_journal_flag_rwsem);
+	percpu_up_write(&sbi->s_writepages_rwsem);
 
 	if (val)
 		up_write(&EXT4_I(inode)->i_mmap_sem);