exfat: change to get file size from DataLength

In stream extension directory entry, the ValidDataLength
field describes how far into the data stream user data has
been written, and the DataLength field describes the file
size.

Signed-off-by: Yuezhang Mo <[email protected]>
Reviewed-by: Andy Wu <[email protected]>
Reviewed-by: Aoyama Wataru <[email protected]>
Reviewed-by: Sungjong Seo <[email protected]>
Signed-off-by: Namjae Jeon <[email protected]>

Author: YuezhangMo

fs/exfat/exfat_fs.h

@@ -207,6 +207,7 @@ struct exfat_dir_entry {
 	unsigned char flags;
 	unsigned short attr;
 	loff_t size;
+	loff_t valid_size;
 	unsigned int num_subdirs;
 	struct timespec64 atime;
 	struct timespec64 mtime;
@@ -316,6 +317,7 @@ struct exfat_inode_info {
 	loff_t i_size_aligned;
 	/* on-disk position of directory entry or 0 */
 	loff_t i_pos;
+	loff_t valid_size;
 	/* hash by i_location */
 	struct hlist_node i_hash_fat;
 	/* protect bmap against truncate */

fs/exfat/file.c

@@ -11,6 +11,7 @@
 #include <linux/fsnotify.h>
 #include <linux/security.h>
 #include <linux/msdos_fs.h>
+#include <linux/writeback.h>
 
 #include "exfat_raw.h"
 #include "exfat_fs.h"
@@ -26,6 +27,7 @@ static int exfat_cont_expand(struct inode *inode, loff_t size)
 		return err;
 
 	inode_set_mtime_to_ts(inode, inode_set_ctime_current(inode));
+	EXFAT_I(inode)->valid_size = size;
 	mark_inode_dirty(inode);
 
 	if (!IS_SYNC(inode))
@@ -146,6 +148,9 @@ int __exfat_truncate(struct inode *inode)
 		ei->start_clu = EXFAT_EOF_CLUSTER;
 	}
 
+	if (i_size_read(inode) < ei->valid_size)
+		ei->valid_size = i_size_read(inode);
+
 	if (ei->type == TYPE_FILE)
 		ei->attr |= EXFAT_ATTR_ARCHIVE;
 
@@ -474,15 +479,124 @@ int exfat_file_fsync(struct file *filp, loff_t start, loff_t end, int datasync)
 	return blkdev_issue_flush(inode->i_sb->s_bdev);
 }
 
+static int exfat_file_zeroed_range(struct file *file, loff_t start, loff_t end)
+{
+	int err;
+	struct inode *inode = file_inode(file);
+	struct address_space *mapping = inode->i_mapping;
+	const struct address_space_operations *ops = mapping->a_ops;
+
+	while (start < end) {
+		u32 zerofrom, len;
+		struct page *page = NULL;
+
+		zerofrom = start & (PAGE_SIZE - 1);
+		len = PAGE_SIZE - zerofrom;
+		if (start + len > end)
+			len = end - start;
+
+		err = ops->write_begin(file, mapping, start, len, &page, NULL);
+		if (err)
+			goto out;
+
+		zero_user_segment(page, zerofrom, zerofrom + len);
+
+		err = ops->write_end(file, mapping, start, len, len, page, NULL);
+		if (err < 0)
+			goto out;
+		start += len;
+
+		balance_dirty_pages_ratelimited(mapping);
+		cond_resched();
+	}
+
+out:
+	return err;
+}
+
+static ssize_t exfat_file_write_iter(struct kiocb *iocb, struct iov_iter *iter)
+{
+	ssize_t ret;
+	struct file *file = iocb->ki_filp;
+	struct inode *inode = file_inode(file);
+	struct exfat_inode_info *ei = EXFAT_I(inode);
+	loff_t pos = iocb->ki_pos;
+	loff_t valid_size;
+
+	inode_lock(inode);
+
+	valid_size = ei->valid_size;
+
+	ret = generic_write_checks(iocb, iter);
+	if (ret < 0)
+		goto unlock;
+
+	if (pos > valid_size) {
+		ret = exfat_file_zeroed_range(file, valid_size, pos);
+		if (ret < 0 && ret != -ENOSPC) {
+			exfat_err(inode->i_sb,
+				"write: fail to zero from %llu to %llu(%zd)",
+				valid_size, pos, ret);
+		}
+		if (ret < 0)
+			goto unlock;
+	}
+
+	ret = __generic_file_write_iter(iocb, iter);
+	if (ret < 0)
+		goto unlock;
+
+	inode_unlock(inode);
+
+	if (pos > valid_size)
+		pos = valid_size;
+
+	if (iocb_is_dsync(iocb) && iocb->ki_pos > pos) {
+		ssize_t err = vfs_fsync_range(file, pos, iocb->ki_pos - 1,
+				iocb->ki_flags & IOCB_SYNC);
+		if (err < 0)
+			return err;
+	}
+
+	return ret;
+
+unlock:
+	inode_unlock(inode);
+
+	return ret;
+}
+
+static int exfat_file_mmap(struct file *file, struct vm_area_struct *vma)
+{
+	int ret;
+	struct inode *inode = file_inode(file);
+	struct exfat_inode_info *ei = EXFAT_I(inode);
+	loff_t start = ((loff_t)vma->vm_pgoff << PAGE_SHIFT);
+	loff_t end = min_t(loff_t, i_size_read(inode),
+			start + vma->vm_end - vma->vm_start);
+
+	if ((vma->vm_flags & VM_WRITE) && ei->valid_size < end) {
+		ret = exfat_file_zeroed_range(file, ei->valid_size, end);
+		if (ret < 0) {
+			exfat_err(inode->i_sb,
+				  "mmap: fail to zero from %llu to %llu(%d)",
+				  start, end, ret);
+			return ret;
+		}
+	}
+
+	return generic_file_mmap(file, vma);
+}
+
 const struct file_operations exfat_file_operations = {
 	.llseek		= generic_file_llseek,
 	.read_iter	= generic_file_read_iter,
-	.write_iter	= generic_file_write_iter,
+	.write_iter	= exfat_file_write_iter,
 	.unlocked_ioctl = exfat_ioctl,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = exfat_compat_ioctl,
 #endif
-	.mmap		= generic_file_mmap,
+	.mmap		= exfat_file_mmap,
 	.fsync		= exfat_file_fsync,
 	.splice_read	= filemap_splice_read,
 	.splice_write	= iter_file_splice_write,

fs/exfat/inode.c

@@ -75,8 +75,8 @@ int __exfat_write_inode(struct inode *inode, int sync)
 	if (ei->start_clu == EXFAT_EOF_CLUSTER)
 		on_disk_size = 0;
 
-	ep2->dentry.stream.valid_size = cpu_to_le64(on_disk_size);
-	ep2->dentry.stream.size = ep2->dentry.stream.valid_size;
+	ep2->dentry.stream.valid_size = cpu_to_le64(ei->valid_size);
+	ep2->dentry.stream.size = cpu_to_le64(on_disk_size);
 	if (on_disk_size) {
 		ep2->dentry.stream.flags = ei->flags;
 		ep2->dentry.stream.start_clu = cpu_to_le32(ei->start_clu);
@@ -278,6 +278,7 @@ static int exfat_get_block(struct inode *inode, sector_t iblock,
 	unsigned int cluster, sec_offset;
 	sector_t last_block;
 	sector_t phys = 0;
+	sector_t valid_blks;
 	loff_t pos;
 
 	mutex_lock(&sbi->s_lock);
@@ -306,29 +307,88 @@ static int exfat_get_block(struct inode *inode, sector_t iblock,
 	mapped_blocks = sbi->sect_per_clus - sec_offset;
 	max_blocks = min(mapped_blocks, max_blocks);
 
-	/* Treat newly added block / cluster */
-	if (iblock < last_block)
-		create = 0;
-
-	if (create || buffer_delay(bh_result)) {
-		pos = EXFAT_BLK_TO_B((iblock + 1), sb);
+	pos = EXFAT_BLK_TO_B((iblock + 1), sb);
+	if ((create && iblock >= last_block) || buffer_delay(bh_result)) {
 		if (ei->i_size_ondisk < pos)
 			ei->i_size_ondisk = pos;
 	}
 
+	map_bh(bh_result, sb, phys);
+	if (buffer_delay(bh_result))
+		clear_buffer_delay(bh_result);
+
 	if (create) {
+		valid_blks = EXFAT_B_TO_BLK_ROUND_UP(ei->valid_size, sb);
+
+		if (iblock + max_blocks < valid_blks) {
+			/* The range has been written, map it */
+			goto done;
+		} else if (iblock < valid_blks) {
+			/*
+			 * The range has been partially written,
+			 * map the written part.
+			 */
+			max_blocks = valid_blks - iblock;
+			goto done;
+		}
+
+		/* The area has not been written, map and mark as new. */
 		err = exfat_map_new_buffer(ei, bh_result, pos);
 		if (err) {
 			exfat_fs_error(sb,
 					"requested for bmap out of range(pos : (%llu) > i_size_aligned(%llu)\n",
 					pos, ei->i_size_aligned);
 			goto unlock_ret;
 		}
+	} else {
+		valid_blks = EXFAT_B_TO_BLK(ei->valid_size, sb);
+
+		if (iblock + max_blocks < valid_blks) {
+			/* The range has been written, map it */
+			goto done;
+		} else if (iblock < valid_blks) {
+			/*
+			 * The area has been partially written,
+			 * map the written part.
+			 */
+			max_blocks = valid_blks - iblock;
+			goto done;
+		} else if (iblock == valid_blks &&
+			   (ei->valid_size & (sb->s_blocksize - 1))) {
+			/*
+			 * The block has been partially written,
+			 * zero the unwritten part and map the block.
+			 */
+			loff_t size, off;
+
+			max_blocks = 1;
+
+			/*
+			 * For direct read, the unwritten part will be zeroed in
+			 * exfat_direct_IO()
+			 */
+			if (!bh_result->b_folio)
+				goto done;
+
+			pos -= sb->s_blocksize;
+			size = ei->valid_size - pos;
+			off = pos & (PAGE_SIZE - 1);
+
+			folio_set_bh(bh_result, bh_result->b_folio, off);
+			err = bh_read(bh_result, 0);
+			if (err < 0)
+				goto unlock_ret;
+
+			folio_zero_segment(bh_result->b_folio, off + size,
+					off + sb->s_blocksize);
+		} else {
+			/*
+			 * The range has not been written, clear the mapped flag
+			 * to only zero the cache and do not read from disk.
+			 */
+			clear_buffer_mapped(bh_result);
+		}
 	}
-
-	if (buffer_delay(bh_result))
-		clear_buffer_delay(bh_result);
-	map_bh(bh_result, sb, phys);
 done:
 	bh_result->b_size = EXFAT_BLK_TO_B(max_blocks, sb);
 unlock_ret:
@@ -343,6 +403,17 @@ static int exfat_read_folio(struct file *file, struct folio *folio)
 
 static void exfat_readahead(struct readahead_control *rac)
 {
+	struct address_space *mapping = rac->mapping;
+	struct inode *inode = mapping->host;
+	struct exfat_inode_info *ei = EXFAT_I(inode);
+	loff_t pos = readahead_pos(rac);
+
+	/* Range cross valid_size, read it page by page. */
+	if (ei->valid_size < i_size_read(inode) &&
+	    pos <= ei->valid_size &&
+	    ei->valid_size < pos + readahead_length(rac))
+		return;
+
 	mpage_readahead(rac, exfat_get_block);
 }
 
@@ -370,9 +441,7 @@ static int exfat_write_begin(struct file *file, struct address_space *mapping,
 	int ret;
 
 	*pagep = NULL;
-	ret = cont_write_begin(file, mapping, pos, len, pagep, fsdata,
-			       exfat_get_block,
-			       &EXFAT_I(mapping->host)->i_size_ondisk);
+	ret = block_write_begin(mapping, pos, len, pagep, exfat_get_block);
 
 	if (ret < 0)
 		exfat_write_failed(mapping, pos+len);
@@ -400,6 +469,11 @@ static int exfat_write_end(struct file *file, struct address_space *mapping,
 	if (err < len)
 		exfat_write_failed(mapping, pos+len);
 
+	if (!(err < 0) && pos + err > ei->valid_size) {
+		ei->valid_size = pos + err;
+		mark_inode_dirty(inode);
+	}
+
 	if (!(err < 0) && !(ei->attr & EXFAT_ATTR_ARCHIVE)) {
 		inode_set_mtime_to_ts(inode, inode_set_ctime_current(inode));
 		ei->attr |= EXFAT_ATTR_ARCHIVE;
@@ -413,6 +487,8 @@ static ssize_t exfat_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 {
 	struct address_space *mapping = iocb->ki_filp->f_mapping;
 	struct inode *inode = mapping->host;
+	struct exfat_inode_info *ei = EXFAT_I(inode);
+	loff_t pos = iocb->ki_pos;
 	loff_t size = iocb->ki_pos + iov_iter_count(iter);
 	int rw = iov_iter_rw(iter);
 	ssize_t ret;
@@ -436,8 +512,21 @@ static ssize_t exfat_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 	 * condition of exfat_get_block() and ->truncate().
 	 */
 	ret = blockdev_direct_IO(iocb, inode, iter, exfat_get_block);
-	if (ret < 0 && (rw & WRITE))
-		exfat_write_failed(mapping, size);
+	if (ret < 0) {
+		if (rw == WRITE)
+			exfat_write_failed(mapping, size);
+
+		if (ret != -EIOCBQUEUED)
+			return ret;
+	} else
+		size = pos + ret;
+
+	/* zero the unwritten part in the partially written block */
+	if (rw == READ && pos < ei->valid_size && ei->valid_size < size) {
+		iov_iter_revert(iter, size - ei->valid_size);
+		iov_iter_zero(size - ei->valid_size, iter);
+	}
+
 	return ret;
 }
 
@@ -537,6 +626,7 @@ static int exfat_fill_inode(struct inode *inode, struct exfat_dir_entry *info)
 	ei->start_clu = info->start_clu;
 	ei->flags = info->flags;
 	ei->type = info->type;
+	ei->valid_size = info->valid_size;
 
 	ei->version = 0;
 	ei->hint_stat.eidx = 0;