1 files changed, 170 insertions, 164 deletions

diff --git a/fs/erofs/data.c b/fs/erofs/data.c
index 52524bd9698b..6a329c329f43 100644
--- a/fs/erofs/data.c
+++ b/fs/erofs/data.c
@@ -10,10 +10,10 @@
 
 void erofs_unmap_metabuf(struct erofs_buf *buf)
 {
-	if (buf->kmap_type == EROFS_KMAP)
-		kunmap_local(buf->base);
+	if (!buf->base)
+		return;
+	kunmap_local(buf->base);
 	buf->base = NULL;
-	buf->kmap_type = EROFS_NO_KMAP;
 }
 
 void erofs_put_metabuf(struct erofs_buf *buf)
@@ -21,127 +21,91 @@ void erofs_put_metabuf(struct erofs_buf *buf)
 	if (!buf->page)
 		return;
 	erofs_unmap_metabuf(buf);
-	put_page(buf->page);
+	folio_put(page_folio(buf->page));
 	buf->page = NULL;
 }
 
-/*
- * Derive the block size from inode->i_blkbits to make compatible with
- * anonymous inode in fscache mode.
- */
-void *erofs_bread(struct erofs_buf *buf, erofs_blk_t blkaddr,
-		  enum erofs_kmap_type type)
+void *erofs_bread(struct erofs_buf *buf, erofs_off_t offset, bool need_kmap)
 {
-	struct inode *inode = buf->inode;
-	erofs_off_t offset = (erofs_off_t)blkaddr << inode->i_blkbits;
-	pgoff_t index = offset >> PAGE_SHIFT;
-	struct page *page = buf->page;
-	struct folio *folio;
-	unsigned int nofs_flag;
-
-	if (!page || page->index != index) {
-		erofs_put_metabuf(buf);
+	pgoff_t index = (buf->off + offset) >> PAGE_SHIFT;
+	struct folio *folio = NULL;
 
-		nofs_flag = memalloc_nofs_save();
-		folio = read_cache_folio(inode->i_mapping, index, NULL, NULL);
-		memalloc_nofs_restore(nofs_flag);
+	if (buf->page) {
+		folio = page_folio(buf->page);
+		if (folio_file_page(folio, index) != buf->page)
+			erofs_unmap_metabuf(buf);
+	}
+	if (!folio || !folio_contains(folio, index)) {
+		erofs_put_metabuf(buf);
+		folio = read_mapping_folio(buf->mapping, index, buf->file);
 		if (IS_ERR(folio))
 			return folio;
-
-		/* should already be PageUptodate, no need to lock page */
-		page = folio_file_page(folio, index);
-		buf->page = page;
-	}
-	if (buf->kmap_type == EROFS_NO_KMAP) {
-		if (type == EROFS_KMAP)
-			buf->base = kmap_local_page(page);
-		buf->kmap_type = type;
-	} else if (buf->kmap_type != type) {
-		DBG_BUGON(1);
-		return ERR_PTR(-EFAULT);
 	}
-	if (type == EROFS_NO_KMAP)
+	buf->page = folio_file_page(folio, index);
+	if (!need_kmap)
 		return NULL;
+	if (!buf->base)
+		buf->base = kmap_local_page(buf->page);
 	return buf->base + (offset & ~PAGE_MASK);
 }
 
 void erofs_init_metabuf(struct erofs_buf *buf, struct super_block *sb)
 {
-	if (erofs_is_fscache_mode(sb))
-		buf->inode = EROFS_SB(sb)->s_fscache->inode;
+	struct erofs_sb_info *sbi = EROFS_SB(sb);
+
+	buf->file = NULL;
+	buf->off = sbi->dif0.fsoff;
+	if (erofs_is_fileio_mode(sbi)) {
+		buf->file = sbi->dif0.file;	/* some fs like FUSE needs it */
+		buf->mapping = buf->file->f_mapping;
+	} else if (erofs_is_fscache_mode(sb))
+		buf->mapping = sbi->dif0.fscache->inode->i_mapping;
 	else
-		buf->inode = sb->s_bdev->bd_inode;
+		buf->mapping = sb->s_bdev->bd_mapping;
 }
 
 void *erofs_read_metabuf(struct erofs_buf *buf, struct super_block *sb,
-			 erofs_blk_t blkaddr, enum erofs_kmap_type type)
+			 erofs_off_t offset, bool need_kmap)
 {
 	erofs_init_metabuf(buf, sb);
-	return erofs_bread(buf, blkaddr, type);
-}
-
-static int erofs_map_blocks_flatmode(struct inode *inode,
-				     struct erofs_map_blocks *map)
-{
-	erofs_blk_t nblocks, lastblk;
-	u64 offset = map->m_la;
-	struct erofs_inode *vi = EROFS_I(inode);
-	struct super_block *sb = inode->i_sb;
-	bool tailendpacking = (vi->datalayout == EROFS_INODE_FLAT_INLINE);
-
-	nblocks = erofs_iblks(inode);
-	lastblk = nblocks - tailendpacking;
-
-	/* there is no hole in flatmode */
-	map->m_flags = EROFS_MAP_MAPPED;
-	if (offset < erofs_pos(sb, lastblk)) {
-		map->m_pa = erofs_pos(sb, vi->raw_blkaddr) + map->m_la;
-		map->m_plen = erofs_pos(sb, lastblk) - offset;
-	} else if (tailendpacking) {
-		map->m_pa = erofs_iloc(inode) + vi->inode_isize +
-			vi->xattr_isize + erofs_blkoff(sb, offset);
-		map->m_plen = inode->i_size - offset;
-
-		/* inline data should be located in the same meta block */
-		if (erofs_blkoff(sb, map->m_pa) + map->m_plen > sb->s_blocksize) {
-			erofs_err(sb, "inline data cross block boundary @ nid %llu",
-				  vi->nid);
-			DBG_BUGON(1);
-			return -EFSCORRUPTED;
-		}
-		map->m_flags |= EROFS_MAP_META;
-	} else {
-		erofs_err(sb, "internal error @ nid: %llu (size %llu), m_la 0x%llx",
-			  vi->nid, inode->i_size, map->m_la);
-		DBG_BUGON(1);
-		return -EIO;
-	}
-	return 0;
+	return erofs_bread(buf, offset, need_kmap);
 }
 
 int erofs_map_blocks(struct inode *inode, struct erofs_map_blocks *map)
 {
+	struct erofs_buf buf = __EROFS_BUF_INITIALIZER;
 	struct super_block *sb = inode->i_sb;
+	unsigned int unit, blksz = sb->s_blocksize;
 	struct erofs_inode *vi = EROFS_I(inode);
 	struct erofs_inode_chunk_index *idx;
-	struct erofs_buf buf = __EROFS_BUF_INITIALIZER;
-	u64 chunknr;
-	unsigned int unit;
+	erofs_blk_t startblk, addrmask;
+	bool tailpacking;
 	erofs_off_t pos;
-	void *kaddr;
+	u64 chunknr;
 	int err = 0;
 
 	trace_erofs_map_blocks_enter(inode, map, 0);
 	map->m_deviceid = 0;
-	if (map->m_la >= inode->i_size) {
-		/* leave out-of-bound access unmapped */
-		map->m_flags = 0;
-		map->m_plen = 0;
+	map->m_flags = 0;
+	if (map->m_la >= inode->i_size)
 		goto out;
-	}
 
 	if (vi->datalayout != EROFS_INODE_CHUNK_BASED) {
-		err = erofs_map_blocks_flatmode(inode, map);
+		tailpacking = (vi->datalayout == EROFS_INODE_FLAT_INLINE);
+		if (!tailpacking && vi->startblk == EROFS_NULL_ADDR)
+			goto out;
+		pos = erofs_pos(sb, erofs_iblks(inode) - tailpacking);
+
+		map->m_flags = EROFS_MAP_MAPPED;
+		if (map->m_la < pos) {
+			map->m_pa = erofs_pos(sb, vi->startblk) + map->m_la;
+			map->m_llen = pos - map->m_la;
+		} else {
+			map->m_pa = erofs_iloc(inode) + vi->inode_isize +
+				vi->xattr_isize + erofs_blkoff(sb, map->m_la);
+			map->m_llen = inode->i_size - map->m_la;
+			map->m_flags |= EROFS_MAP_META;
+		}
 		goto out;
 	}
 
@@ -154,60 +118,67 @@ int erofs_map_blocks(struct inode *inode, struct erofs_map_blocks *map)
 	pos = ALIGN(erofs_iloc(inode) + vi->inode_isize +
 		    vi->xattr_isize, unit) + unit * chunknr;
 
-	kaddr = erofs_read_metabuf(&buf, sb, erofs_blknr(sb, pos), EROFS_KMAP);
-	if (IS_ERR(kaddr)) {
-		err = PTR_ERR(kaddr);
+	idx = erofs_read_metabuf(&buf, sb, pos, true);
+	if (IS_ERR(idx)) {
+		err = PTR_ERR(idx);
 		goto out;
 	}
 	map->m_la = chunknr << vi->chunkbits;
-	map->m_plen = min_t(erofs_off_t, 1UL << vi->chunkbits,
-			round_up(inode->i_size - map->m_la, sb->s_blocksize));
-
-	/* handle block map */
-	if (!(vi->chunkformat & EROFS_CHUNK_FORMAT_INDEXES)) {
-		__le32 *blkaddr = kaddr + erofs_blkoff(sb, pos);
-
-		if (le32_to_cpu(*blkaddr) == EROFS_NULL_ADDR) {
-			map->m_flags = 0;
-		} else {
-			map->m_pa = erofs_pos(sb, le32_to_cpu(*blkaddr));
+	map->m_llen = min_t(erofs_off_t, 1UL << vi->chunkbits,
+			    round_up(inode->i_size - map->m_la, blksz));
+	if (vi->chunkformat & EROFS_CHUNK_FORMAT_INDEXES) {
+		addrmask = (vi->chunkformat & EROFS_CHUNK_FORMAT_48BIT) ?
+			BIT_ULL(48) - 1 : BIT_ULL(32) - 1;
+		startblk = (((u64)le16_to_cpu(idx->startblk_hi) << 32) |
+			    le32_to_cpu(idx->startblk_lo)) & addrmask;
+		if ((startblk ^ EROFS_NULL_ADDR) & addrmask) {
+			map->m_deviceid = le16_to_cpu(idx->device_id) &
+				EROFS_SB(sb)->device_id_mask;
+			map->m_pa = erofs_pos(sb, startblk);
+			map->m_flags = EROFS_MAP_MAPPED;
+		}
+	} else {
+		startblk = le32_to_cpu(*(__le32 *)idx);
+		if (startblk != (u32)EROFS_NULL_ADDR) {
+			map->m_pa = erofs_pos(sb, startblk);
 			map->m_flags = EROFS_MAP_MAPPED;
 		}
-		goto out_unlock;
-	}
-	/* parse chunk indexes */
-	idx = kaddr + erofs_blkoff(sb, pos);
-	switch (le32_to_cpu(idx->blkaddr)) {
-	case EROFS_NULL_ADDR:
-		map->m_flags = 0;
-		break;
-	default:
-		map->m_deviceid = le16_to_cpu(idx->device_id) &
-			EROFS_SB(sb)->device_id_mask;
-		map->m_pa = erofs_pos(sb, le32_to_cpu(idx->blkaddr));
-		map->m_flags = EROFS_MAP_MAPPED;
-		break;
 	}
-out_unlock:
 	erofs_put_metabuf(&buf);
 out:
-	if (!err)
-		map->m_llen = map->m_plen;
+	if (!err) {
+		map->m_plen = map->m_llen;
+		/* inline data should be located in the same meta block */
+		if ((map->m_flags & EROFS_MAP_META) &&
+		    erofs_blkoff(sb, map->m_pa) + map->m_plen > blksz) {
+			erofs_err(sb, "inline data across blocks @ nid %llu", vi->nid);
+			DBG_BUGON(1);
+			return -EFSCORRUPTED;
+		}
+	}
 	trace_erofs_map_blocks_exit(inode, map, 0, err);
 	return err;
 }
 
+static void erofs_fill_from_devinfo(struct erofs_map_dev *map,
+		struct super_block *sb, struct erofs_device_info *dif)
+{
+	map->m_sb = sb;
+	map->m_dif = dif;
+	map->m_bdev = NULL;
+	if (dif->file && S_ISBLK(file_inode(dif->file)->i_mode))
+		map->m_bdev = file_bdev(dif->file);
+}
+
 int erofs_map_dev(struct super_block *sb, struct erofs_map_dev *map)
 {
 	struct erofs_dev_context *devs = EROFS_SB(sb)->devs;
 	struct erofs_device_info *dif;
+	erofs_off_t startoff;
 	int id;
 
-	map->m_bdev = sb->s_bdev;
-	map->m_daxdev = EROFS_SB(sb)->dax_dev;
-	map->m_dax_part_off = EROFS_SB(sb)->dax_part_off;
-	map->m_fscache = EROFS_SB(sb)->s_fscache;
-
+	erofs_fill_from_devinfo(map, sb, &EROFS_SB(sb)->dif0);
+	map->m_bdev = sb->s_bdev;	/* use s_bdev for the primary device */
 	if (map->m_deviceid) {
 		down_read(&devs->rwsem);
 		dif = idr_find(&devs->tree, map->m_deviceid - 1);
@@ -216,33 +187,23 @@ int erofs_map_dev(struct super_block *sb, struct erofs_map_dev *map)
 			return -ENODEV;
 		}
 		if (devs->flatdev) {
-			map->m_pa += erofs_pos(sb, dif->mapped_blkaddr);
+			map->m_pa += erofs_pos(sb, dif->uniaddr);
 			up_read(&devs->rwsem);
 			return 0;
 		}
-		map->m_bdev = dif->bdev_file ? file_bdev(dif->bdev_file) : NULL;
-		map->m_daxdev = dif->dax_dev;
-		map->m_dax_part_off = dif->dax_part_off;
-		map->m_fscache = dif->fscache;
+		erofs_fill_from_devinfo(map, sb, dif);
 		up_read(&devs->rwsem);
 	} else if (devs->extra_devices && !devs->flatdev) {
 		down_read(&devs->rwsem);
 		idr_for_each_entry(&devs->tree, dif, id) {
-			erofs_off_t startoff, length;
-
-			if (!dif->mapped_blkaddr)
+			if (!dif->uniaddr)
 				continue;
-			startoff = erofs_pos(sb, dif->mapped_blkaddr);
-			length = erofs_pos(sb, dif->blocks);
 
+			startoff = erofs_pos(sb, dif->uniaddr);
 			if (map->m_pa >= startoff &&
-			    map->m_pa < startoff + length) {
+			    map->m_pa < startoff + erofs_pos(sb, dif->blocks)) {
 				map->m_pa -= startoff;
-				map->m_bdev = dif->bdev_file ?
-					      file_bdev(dif->bdev_file) : NULL;
-				map->m_daxdev = dif->dax_dev;
-				map->m_dax_part_off = dif->dax_part_off;
-				map->m_fscache = dif->fscache;
+				erofs_fill_from_devinfo(map, sb, dif);
 				break;
 			}
 		}
@@ -251,6 +212,42 @@ int erofs_map_dev(struct super_block *sb, struct erofs_map_dev *map)
 	return 0;
 }
 
+/*
+ * bit 30: I/O error occurred on this folio
+ * bit 0 - 29: remaining parts to complete this folio
+ */
+#define EROFS_ONLINEFOLIO_EIO			(1 << 30)
+
+void erofs_onlinefolio_init(struct folio *folio)
+{
+	union {
+		atomic_t o;
+		void *v;
+	} u = { .o = ATOMIC_INIT(1) };
+
+	folio->private = u.v;	/* valid only if file-backed folio is locked */
+}
+
+void erofs_onlinefolio_split(struct folio *folio)
+{
+	atomic_inc((atomic_t *)&folio->private);
+}
+
+void erofs_onlinefolio_end(struct folio *folio, int err)
+{
+	int orig, v;
+
+	do {
+		orig = atomic_read((atomic_t *)&folio->private);
+		v = (orig - 1) | (err ? EROFS_ONLINEFOLIO_EIO : 0);
+	} while (atomic_cmpxchg((atomic_t *)&folio->private, orig, v) != orig);
+
+	if (v & ~EROFS_ONLINEFOLIO_EIO)
+		return;
+	folio->private = 0;
+	folio_end_read(folio, !(v & EROFS_ONLINEFOLIO_EIO));
+}
+
 static int erofs_iomap_begin(struct inode *inode, loff_t offset, loff_t length,
 		unsigned int flags, struct iomap *iomap, struct iomap *srcmap)
 {
@@ -276,7 +273,7 @@ static int erofs_iomap_begin(struct inode *inode, loff_t offset, loff_t length,
 
 	iomap->offset = map.m_la;
 	if (flags & IOMAP_DAX)
-		iomap->dax_dev = mdev.m_daxdev;
+		iomap->dax_dev = mdev.m_dif->dax_dev;
 	else
 		iomap->bdev = mdev.m_bdev;
 	iomap->length = map.m_llen;
@@ -296,17 +293,16 @@ static int erofs_iomap_begin(struct inode *inode, loff_t offset, loff_t length,
 		struct erofs_buf buf = __EROFS_BUF_INITIALIZER;
 
 		iomap->type = IOMAP_INLINE;
-		ptr = erofs_read_metabuf(&buf, sb,
-				erofs_blknr(sb, mdev.m_pa), EROFS_KMAP);
+		ptr = erofs_read_metabuf(&buf, sb, mdev.m_pa, true);
 		if (IS_ERR(ptr))
 			return PTR_ERR(ptr);
-		iomap->inline_data = ptr + erofs_blkoff(sb, mdev.m_pa);
+		iomap->inline_data = ptr;
 		iomap->private = buf.base;
 	} else {
 		iomap->type = IOMAP_MAPPED;
-		iomap->addr = mdev.m_pa;
+		iomap->addr = mdev.m_dif->fsoff + mdev.m_pa;
 		if (flags & IOMAP_DAX)
-			iomap->addr += mdev.m_dax_part_off;
+			iomap->addr += mdev.m_dif->dax_part_off;
 	}
 	return 0;
 }
@@ -320,7 +316,6 @@ static int erofs_iomap_end(struct inode *inode, loff_t pos, loff_t length,
 		struct erofs_buf buf = {
 			.page = kmap_to_page(ptr),
 			.base = ptr,
-			.kmap_type = EROFS_KMAP,
 		};
 
 		DBG_BUGON(iomap->type != IOMAP_INLINE);
@@ -381,27 +376,14 @@ static ssize_t erofs_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
 	if (IS_DAX(inode))
 		return dax_iomap_rw(iocb, to, &erofs_iomap_ops);
 #endif
-	if (iocb->ki_flags & IOCB_DIRECT) {
-		struct block_device *bdev = inode->i_sb->s_bdev;
-		unsigned int blksize_mask;
-
-		if (bdev)
-			blksize_mask = bdev_logical_block_size(bdev) - 1;
-		else
-			blksize_mask = i_blocksize(inode) - 1;
-
-		if ((iocb->ki_pos | iov_iter_count(to) |
-		     iov_iter_alignment(to)) & blksize_mask)
-			return -EINVAL;
-
+	if ((iocb->ki_flags & IOCB_DIRECT) && inode->i_sb->s_bdev)
 		return iomap_dio_rw(iocb, to, &erofs_iomap_ops,
 				    NULL, 0, NULL, 0);
-	}
 	return filemap_read(iocb, to, 0);
 }
 
 /* for uncompressed (aligned) files and raw access for other files */
-const struct address_space_operations erofs_raw_access_aops = {
+const struct address_space_operations erofs_aops = {
 	.read_folio = erofs_read_folio,
 	.readahead = erofs_readahead,
 	.bmap = erofs_bmap,
@@ -443,8 +425,32 @@ static int erofs_file_mmap(struct file *file, struct vm_area_struct *vma)
 #define erofs_file_mmap	generic_file_readonly_mmap
 #endif
 
+static loff_t erofs_file_llseek(struct file *file, loff_t offset, int whence)
+{
+	struct inode *inode = file->f_mapping->host;
+	const struct iomap_ops *ops = &erofs_iomap_ops;
+
+	if (erofs_inode_is_data_compressed(EROFS_I(inode)->datalayout))
+#ifdef CONFIG_EROFS_FS_ZIP
+		ops = &z_erofs_iomap_report_ops;
+#else
+		return generic_file_llseek(file, offset, whence);
+#endif
+
+	if (whence == SEEK_HOLE)
+		offset = iomap_seek_hole(inode, offset, ops);
+	else if (whence == SEEK_DATA)
+		offset = iomap_seek_data(inode, offset, ops);
+	else
+		return generic_file_llseek(file, offset, whence);
+
+	if (offset < 0)
+		return offset;
+	return vfs_setpos(file, offset, inode->i_sb->s_maxbytes);
+}
+
 const struct file_operations erofs_file_fops = {
-	.llseek		= generic_file_llseek,
+	.llseek		= erofs_file_llseek,
 	.read_iter	= erofs_file_read_iter,
 	.mmap		= erofs_file_mmap,
 	.get_unmapped_area = thp_get_unmapped_area,