Merge tag 'erofs-for-7.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/xiang/erofs

Pull erofs updates from Gao Xiang: "The most notable enhancement is to allow passing source fds via fsconfig() for composefs. The others are all various fixes: - Allow source fds via fsconfig(), in addition to source paths - Use dedicated metadata inodes for file-backed mounts - Disallow invalid interlaced ztailpacking pclusters - Validate on-disk compression algorithm IDs against supported ones - Fix unused pcluster pools on higher page-size platforms" * tag 'erofs-for-7.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/xiang/erofs: erofs: fix unused pcluster_pools for higher page sizes erofs: guard on-disk algorithm IDs against Z_EROFS_COMPRESSION_MAX erofs: fix interlaced ztailpacking pclusters erofs: use dedicated meta inodes for file-backed mounts erofs: accept source file descriptor via fsconfig erofs: fix typo in error messages
2026-08-30 11:03:07 -04:00 · 2026-08-20 12:24:48 -07:00
parent 62abb6cffd 988c3c186c
commit 8f421dfec2
8 changed files with 182 additions and 74 deletions
--- a/Documentation/filesystems/erofs.rst
+++ b/Documentation/filesystems/erofs.rst
@@ -137,8 +137,25 @@ fsoffset=%llu          Specify block-aligned filesystem offset for the primary d
 inode_share            Enable inode page sharing for this filesystem.  Inodes with
                       identical content within the same domain ID can share the
                       page cache.
+source=%s              (For file-backed mounts) Specify the backing image as a path
+                       or as an already-opened file descriptor.
 ===================    =========================================================

+File-backed mounts
+==================
+
+When CONFIG_EROFS_FS_BACKED_BY_FILE is enabled, EROFS file-backed images
+can be mounted directly without a loopback block device.  The backing file
+can be given either as a path, or as an already-opened file descriptor.
+
+When a file descriptor is used, the kernel resolves its path and records it
+so that /proc/mounts and similar interfaces can still report the mount
+source.
+
+Only regular files are accepted as backing files; to mount an image that
+resides on a block device, use the traditional block device mount path
+instead.
+
 Sysfs Entries
 =============

--- a/fs/erofs/data.c
+++ b/fs/erofs/data.c
@@ -30,20 +30,6 @@ void *erofs_bread(struct erofs_buf *buf, erofs_off_t offset, bool need_kmap)
 {
 	pgoff_t index = (buf->off + offset) >> PAGE_SHIFT;
 	struct folio *folio = NULL;
-	loff_t fpos;
-	int err;
-
-	/*
-	 * Metadata access for file-backed mounts reuses page cache of backing
-	 * fs inodes (only folio data will be needed) to prevent double caching.
-	 * However, the data access range must be verified here in advance.
-	 */
-	if (buf->file) {
-		fpos = (loff_t)index << PAGE_SHIFT;
-		err = rw_verify_area(READ, buf->file, &fpos, PAGE_SIZE);
-		if (err < 0)
-			return ERR_PTR(err);
-	}

 	if (buf->page) {
 		folio = page_folio(buf->page);
@@ -52,7 +38,8 @@ void *erofs_bread(struct erofs_buf *buf, erofs_off_t offset, bool need_kmap)
 	}
 	if (!folio || !folio_contains(folio, index)) {
 		erofs_put_metabuf(buf);
-		folio = read_mapping_folio(buf->mapping, index, buf->file);
+		folio = read_cache_folio(buf->mapping, index,
+				buf->mc ? erofs_read_meta_folio : NULL, NULL);
 		if (IS_ERR(folio))
 			return folio;
 	}
@@ -69,19 +56,20 @@ int erofs_init_metabuf(struct erofs_buf *buf, struct super_block *sb,
 {
 	struct erofs_sb_info *sbi = EROFS_SB(sb);

-	buf->file = NULL;
+	buf->mc = false;
 	if (in_metabox) {
 		if (unlikely(!sbi->metabox_inode))
 			return -EFSCORRUPTED;
 		buf->mapping = sbi->metabox_inode->i_mapping;
 		return 0;
 	}
-	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
+		buf->mapping = sbi->managed_cache->i_mapping;
+		buf->mc = true;
+	} else {
+		buf->off = sbi->dif0.fsoff;
 		buf->mapping = sb->s_bdev->bd_mapping;
+	}
 	return 0;
 }

--- a/fs/erofs/decompressor.c
+++ b/fs/erofs/decompressor.c
@@ -299,6 +299,7 @@ static const char *z_erofs_transform_plain(struct z_erofs_decompress_req *rq,
 		return ERR_PTR(-EOPNOTSUPP);
 	if (rq->alg == Z_EROFS_COMPRESSION_INTERLACED) {
 		cur = bs - (rq->pageofs_out & (bs - 1));
+		DBG_BUGON(rq->pageofs_in & (bs - 1));
 		pi = (rq->pageofs_in + rq->inputsize - cur) & ~PAGE_MASK;
 		cur = min(cur, rq->outputsize);
 		if (cur && rq->out[0]) {
--- a/fs/erofs/fileio.c
+++ b/fs/erofs/fileio.c
@@ -194,3 +194,24 @@ const struct address_space_operations erofs_fileio_aops = {
 	.read_folio = erofs_fileio_read_folio,
 	.readahead = erofs_fileio_readahead,
 };
+
+int erofs_read_meta_folio(struct file *file, struct folio *folio)
+{
+	struct erofs_fileio io = {
+		.dev = { .m_pa = folio_pos(folio), },
+	};
+	struct inode *inode = folio_inode(folio);
+	int err;
+
+	err = erofs_map_dev(inode->i_sb, &io.dev);
+	if (err)
+		return err;
+
+	io.rq = erofs_fileio_rq_alloc(&io.dev);
+	io.rq->bio.bi_iter.bi_sector =
+		(io.dev.m_dif->fsoff + io.dev.m_pa) >> 9;
+	erofs_onlinefolio_init(folio);
+	bio_add_folio_nofail(&io.rq->bio, folio, folio_size(folio), 0);
+	erofs_fileio_rq_submit(io.rq);
+	return 0;
+}
--- a/fs/erofs/internal.h
+++ b/fs/erofs/internal.h
@@ -98,11 +98,9 @@ struct erofs_sb_info {
 	unsigned int sync_decompress;	/* strategy for sync decompression */
 	unsigned int shrinker_run_no;

-	/* pseudo inode to manage cached pages */
-	struct inode *managed_cache;
-
 	struct erofs_sb_lz4_info lz4;
 #endif	/* CONFIG_EROFS_FS_ZIP */
+	struct inode *managed_cache; /* pseudo inode to cache physical data */
 	struct inode *packed_inode;
 	struct inode *metabox_inode;
 	struct erofs_dev_context *devs;
@@ -176,10 +174,10 @@ enum {

 struct erofs_buf {
 	struct address_space *mapping;
-	struct file *file;
 	u64 off;
 	struct page *page;
 	void *base;
+	bool mc;
 };
 #define __EROFS_BUF_INITIALIZER	((struct erofs_buf){ .page = NULL })

@@ -269,7 +267,7 @@ struct erofs_inode {
 #ifdef CONFIG_EROFS_FS_ZIP
 		struct {
 			unsigned short z_advise;
-			unsigned char  z_algorithmtype[2];
+			unsigned char  z_algofmt[2];
 			unsigned char  z_lclusterbits;
 			union {
 				u64    z_tailextent_headlcn;
@@ -399,6 +397,12 @@ extern const struct file_operations erofs_ishare_fops;

 extern const struct iomap_ops z_erofs_iomap_report_ops;

+int erofs_setup_managed_cache(struct super_block *sb);
+#ifdef CONFIG_EROFS_FS_BACKED_BY_FILE
+int erofs_read_meta_folio(struct file *file, struct folio *folio);
+#else
+#define erofs_read_meta_folio NULL
+#endif
 void *erofs_read_metadata(struct super_block *sb, struct erofs_buf *buf,
 			  erofs_off_t *offset, int *lengthp);
 void erofs_unmap_metabuf(struct erofs_buf *buf);
--- a/fs/erofs/super.c
+++ b/fs/erofs/super.c
@@ -386,6 +386,7 @@ static void erofs_default_options(struct erofs_sb_info *sbi)
 enum {
 	Opt_user_xattr, Opt_acl, Opt_cache_strategy, Opt_dax, Opt_dax_enum,
 	Opt_device, Opt_domain_id, Opt_directio, Opt_fsoffset, Opt_inode_share,
+	Opt_source,
 };

 static const struct constant_table erofs_param_cache_strategy[] = {
@@ -402,17 +403,18 @@ static const struct constant_table erofs_dax_param_enums[] = {
 };

 static const struct fs_parameter_spec erofs_fs_parameters[] = {
-	fsparam_flag_no("user_xattr",	Opt_user_xattr),
-	fsparam_flag_no("acl",		Opt_acl),
-	fsparam_enum("cache_strategy",	Opt_cache_strategy,
+	fsparam_flag_no("user_xattr",		Opt_user_xattr),
+	fsparam_flag_no("acl",			Opt_acl),
+	fsparam_enum("cache_strategy",		Opt_cache_strategy,
 		     erofs_param_cache_strategy),
-	fsparam_flag("dax",             Opt_dax),
-	fsparam_enum("dax",		Opt_dax_enum, erofs_dax_param_enums),
-	fsparam_string("device",	Opt_device),
-	fsparam_string("domain_id",	Opt_domain_id),
-	fsparam_flag_no("directio",	Opt_directio),
-	fsparam_u64("fsoffset",		Opt_fsoffset),
-	fsparam_flag("inode_share",	Opt_inode_share),
+	fsparam_flag("dax",			Opt_dax),
+	fsparam_enum("dax",			Opt_dax_enum, erofs_dax_param_enums),
+	fsparam_string("device",		Opt_device),
+	fsparam_string("domain_id",		Opt_domain_id),
+	fsparam_flag_no("directio",		Opt_directio),
+	fsparam_u64("fsoffset",			Opt_fsoffset),
+	fsparam_flag("inode_share",		Opt_inode_share),
+	fsparam_file_or_string("source",	Opt_source),
 	{}
 };

@@ -437,6 +439,40 @@ static bool erofs_fc_set_dax_mode(struct fs_context *fc, unsigned int mode)
 	return false;
 }

+static int erofs_fc_parse_source(struct fs_context *fc,
+				 struct fs_parameter *param)
+{
+	struct erofs_sb_info *sbi = fc->s_fs_info;
+
+	if (fc->source || sbi->dif0.file)
+		return invalf(fc, "Multiple sources");
+
+	switch (param->type) {
+	case fs_value_is_string:
+		fc->source = param->string;
+		param->string = NULL;
+		return 0;
+	case fs_value_is_file: {
+		char *buf __free(kfree) = kmalloc(PATH_MAX, GFP_KERNEL);
+		char *p;
+
+		if (!buf)
+			return -ENOMEM;
+		p = file_path(param->file, buf, PATH_MAX);
+		if (IS_ERR(p))
+			return PTR_ERR(p);
+		fc->source = kstrdup(p, GFP_KERNEL);
+		if (!fc->source)
+			return -ENOMEM;
+		sbi->dif0.file = no_free_ptr(param->file);
+		return 0;
+	}
+	default:
+		WARN_ON_ONCE(true);
+		return -EINVAL;
+	}
+}
+
 static int erofs_fc_parse_param(struct fs_context *fc,
 				struct fs_parameter *param)
 {
@@ -524,6 +560,8 @@ static int erofs_fc_parse_param(struct fs_context *fc,
 		else
 			set_opt(&sbi->opt, INODE_SHARE);
 		break;
+	case Opt_source:
+		return erofs_fc_parse_source(fc, param);
 	}
 	return 0;
 }
@@ -595,6 +633,21 @@ static const struct export_operations erofs_export_ops = {
 	.get_parent = erofs_get_parent,
 };

+int erofs_setup_managed_cache(struct super_block *sb)
+{
+	if (!EROFS_SB(sb)->managed_cache) {
+		struct inode *inode = new_inode(sb);
+
+		if (!inode)
+			return -ENOMEM;
+		set_nlink(inode, 1);
+		inode->i_size = OFFSET_MAX;
+		mapping_set_gfp_mask(inode->i_mapping, GFP_KERNEL);
+		EROFS_SB(sb)->managed_cache = inode;
+	}
+	return 0;
+}
+
 static int erofs_fc_fill_super(struct super_block *sb, struct fs_context *fc)
 {
 	struct inode *inode;
@@ -607,16 +660,16 @@ static int erofs_fc_fill_super(struct super_block *sb, struct fs_context *fc)
 	sb->s_op = &erofs_sops;

 	if (!sbi->domain_id && test_opt(&sbi->opt, INODE_SHARE)) {
-		errorfc(fc, "domain_id is needed when inode_ishare is on");
+		errorfc(fc, "domain_id is needed when inode_share is on");
 		return -EINVAL;
 	}
 	if (test_opt(&sbi->opt, DAX_ALWAYS) && test_opt(&sbi->opt, INODE_SHARE)) {
-		errorfc(fc, "FSDAX is not allowed when inode_ishare is on");
+		errorfc(fc, "FSDAX is not allowed when inode_share is on");
 		return -EINVAL;
 	}

 	sbi->blkszbits = PAGE_SHIFT;
-	if (!sb->s_bdev) {
+	if (erofs_is_fileio_mode(sbi)) {
 		/*
 		 * (File-backed mounts) EROFS claims it's safe to nest other
 		 * fs contexts (including its own) due to self-controlled RO
@@ -631,19 +684,19 @@ static int erofs_fc_fill_super(struct super_block *sb, struct fs_context *fc)
 		 * It MUST change if another fs plans to support them, which
 		 * may also require adjusting FILESYSTEM_MAX_STACK_DEPTH.
 		 */
-		if (erofs_is_fileio_mode(sbi)) {
-			inode = file_inode(sbi->dif0.file);
-			if ((inode->i_sb->s_op == &erofs_sops &&
-			     !inode->i_sb->s_bdev) ||
-			    inode->i_sb->s_stack_depth) {
-				erofs_err(sb, "file-backed mounts cannot be applied to stacked fses");
-				return -ENOTBLK;
-			}
+		inode = file_inode(sbi->dif0.file);
+		if ((inode->i_sb->s_op == &erofs_sops &&
+		     !inode->i_sb->s_bdev) || inode->i_sb->s_stack_depth) {
+			erofs_err(sb, "file-backed mounts cannot be applied to stacked fses");
+			return -ENOTBLK;
 		}
 		sb->s_blocksize = PAGE_SIZE;
 		sb->s_blocksize_bits = PAGE_SHIFT;

 		err = super_setup_bdi(sb);
+		if (err)
+			return err;
+		err = erofs_setup_managed_cache(sb);
 		if (err)
 			return err;

@@ -743,13 +796,26 @@ static int erofs_fc_fill_super(struct super_block *sb, struct fs_context *fc)

 static int erofs_fc_get_tree(struct fs_context *fc)
 {
+	struct erofs_sb_info *sbi = fc->s_fs_info;
 	int ret;

+	if (sbi->dif0.file) {
+		if (!IS_ENABLED(CONFIG_EROFS_FS_BACKED_BY_FILE)) {
+			errorfc(fc, "source fd option not supported");
+			return -EINVAL;
+		}
+		if (!S_ISREG(file_inode(sbi->dif0.file)->i_mode) ||
+		    !sbi->dif0.file->f_mapping->a_ops->read_folio) {
+			errorfc(fc, "source is unsupported");
+			return -EINVAL;
+		}
+		return get_tree_nodev(fc, erofs_fc_fill_super);
+	}
+
 	ret = get_tree_bdev_flags(fc, erofs_fc_fill_super,
 		IS_ENABLED(CONFIG_EROFS_FS_BACKED_BY_FILE) ?
 			GET_TREE_BDEV_QUIET_LOOKUP : 0);
 	if (IS_ENABLED(CONFIG_EROFS_FS_BACKED_BY_FILE) && ret == -ENOTBLK) {
-		struct erofs_sb_info *sbi = fc->s_fs_info;
 		struct file *file;

 		if (!fc->source)
@@ -873,10 +939,8 @@ static void erofs_drop_internal_inodes(struct erofs_sb_info *sbi)
 	sbi->packed_inode = NULL;
 	iput(sbi->metabox_inode);
 	sbi->metabox_inode = NULL;
-#ifdef CONFIG_EROFS_FS_ZIP
 	iput(sbi->managed_cache);
 	sbi->managed_cache = NULL;
-#endif
 }

 static void erofs_kill_sb(struct super_block *sb)
--- a/fs/erofs/zdata.c
+++ b/fs/erofs/zdata.c
@@ -128,7 +128,17 @@ struct z_erofs_pcluster_slab {
 #define _PCLP(n) { .maxpages = n }

 static struct z_erofs_pcluster_slab pcluster_pool[] __read_mostly = {
-	_PCLP(1), _PCLP(4), _PCLP(16), _PCLP(64), _PCLP(128),
+	_PCLP(1),
+	_PCLP(4),
+#if Z_EROFS_PCLUSTER_MAX_PAGES > 16
+	_PCLP(16),
+#endif
+#if Z_EROFS_PCLUSTER_MAX_PAGES > 64
+	_PCLP(64),
+#endif
+#if Z_EROFS_PCLUSTER_MAX_PAGES > 128
+	_PCLP(128),
+#endif
 	_PCLP(Z_EROFS_PCLUSTER_MAX_PAGES + 1)
 };

@@ -666,21 +676,16 @@ static const struct address_space_operations z_erofs_cache_aops = {

 int z_erofs_init_super(struct super_block *sb)
 {
-	struct inode *inode;
 	int err;

 	err = z_erofs_init_pcpu_workers(sb);
 	if (err)
 		return err;

-	inode = new_inode(sb);
-	if (!inode)
-		return -ENOMEM;
-	set_nlink(inode, 1);
-	inode->i_size = OFFSET_MAX;
-	inode->i_mapping->a_ops = &z_erofs_cache_aops;
-	mapping_set_gfp_mask(inode->i_mapping, GFP_KERNEL);
-	EROFS_SB(sb)->managed_cache = inode;
+	err = erofs_setup_managed_cache(sb);
+	if (err)
+		return err;
+	EROFS_SB(sb)->managed_cache->i_mapping->a_ops = &z_erofs_cache_aops;
 	xa_init(&EROFS_SB(sb)->managed_pslots);
 	return 0;
 }
--- a/fs/erofs/zmap.c
+++ b/fs/erofs/zmap.c
@@ -483,14 +483,14 @@ static int z_erofs_map_blocks_fo(struct inode *inode,
 	}

 	if (m.headtype == Z_EROFS_LCLUSTER_TYPE_PLAIN) {
-		if (vi->z_advise & Z_EROFS_ADVISE_INTERLACED_PCLUSTER)
+		if ((vi->z_advise & Z_EROFS_ADVISE_INTERLACED_PCLUSTER) &&
+		    !(map->m_flags & EROFS_MAP_META))
 			map->m_algorithmformat = Z_EROFS_COMPRESSION_INTERLACED;
 		else
 			map->m_algorithmformat = Z_EROFS_COMPRESSION_SHIFTED;
-	} else if (m.headtype == Z_EROFS_LCLUSTER_TYPE_HEAD2) {
-		map->m_algorithmformat = vi->z_algorithmtype[1];
 	} else {
-		map->m_algorithmformat = vi->z_algorithmtype[0];
+		map->m_algorithmformat =
+			vi->z_algofmt[m.headtype == Z_EROFS_LCLUSTER_TYPE_HEAD2];
 	}

 	if ((flags & EROFS_GET_BLOCKS_FIEMAP) ||
@@ -604,9 +604,14 @@ static int z_erofs_map_blocks_ext(struct inode *inode,
 			if (map->m_plen & Z_EROFS_EXTENT_PLEN_PARTIAL)
 				map->m_flags |= EROFS_MAP_PARTIAL_REF;
 			map->m_plen &= Z_EROFS_EXTENT_PLEN_MASK;
-			if (fmt)
-				map->m_algorithmformat = fmt - 1;
-			else if (interlaced && !((map->m_pa | map->m_plen) & bmask))
+			if (fmt) {
+				map->m_algorithmformat = --fmt;
+				if (fmt >= Z_EROFS_COMPRESSION_MAX) {
+					erofs_err(sb, "unknown algorithm %d @ pos %llu for nid %llu, please upgrade kernel",
+						  fmt, map->m_la, vi->nid);
+					return -EOPNOTSUPP;
+				}
+			} else if (interlaced && !((map->m_pa | map->m_plen) & bmask))
 				map->m_algorithmformat =
 					Z_EROFS_COMPRESSION_INTERLACED;
 			else
@@ -624,7 +629,7 @@ static int z_erofs_fill_inode(struct inode *inode, struct erofs_map_blocks *map)
 	struct super_block *const sb = inode->i_sb;
 	struct z_erofs_map_header *h;
 	erofs_off_t pos;
-	int err = 0;
+	int err = 0, nr;

 	if (test_bit(EROFS_I_Z_INITED_BIT, &vi->flags)) {
 		/*
@@ -667,12 +672,19 @@ static int z_erofs_fill_inode(struct inode *inode, struct erofs_map_blocks *map)
 		goto done;
 	}

-	vi->z_algorithmtype[0] = h->h_algorithmtype & 15;
-	vi->z_algorithmtype[1] = h->h_algorithmtype >> 4;
 	if (vi->z_advise & Z_EROFS_ADVISE_FRAGMENT_PCLUSTER)
 		vi->z_fragmentoff = le32_to_cpu(h->h_fragmentoff);
 	else if (vi->z_advise & Z_EROFS_ADVISE_INLINE_PCLUSTER)
 		vi->z_idata_size = le16_to_cpu(h->h_idata_size);
+	for (nr = 0; nr < 2; ++nr) {
+		vi->z_algofmt[nr] = (h->h_algorithmtype >> (4 * nr)) & 15;
+		if (vi->z_algofmt[nr] >= Z_EROFS_COMPRESSION_MAX) {
+			erofs_err(sb, "unknown HEAD%u format %u for nid %llu, please upgrade kernel",
+				  nr + 1, vi->z_algofmt[nr], vi->nid);
+			err = -EOPNOTSUPP;
+			goto out_unlock;
+		}
+	}

 	if (!erofs_sb_has_big_pcluster(EROFS_SB(sb)) &&
 	    vi->z_advise & (Z_EROFS_ADVISE_BIG_PCLUSTER_1 |
@@ -720,12 +732,8 @@ static int z_erofs_map_sanity_check(struct inode *inode,

 	if (!(map->m_flags & EROFS_MAP_MAPPED))
 		return 0;
-	if (unlikely(map->m_algorithmformat >= Z_EROFS_COMPRESSION_RUNTIME_MAX)) {
-		erofs_err(inode->i_sb, "unknown algorithm %d @ pos %llu for nid %llu, please upgrade kernel",
-			  map->m_algorithmformat, map->m_la, EROFS_I(inode)->nid);
-		return -EOPNOTSUPP;
-	}

+	DBG_BUGON(map->m_algorithmformat >= Z_EROFS_COMPRESSION_RUNTIME_MAX);
 	if (map->m_algorithmformat < Z_EROFS_COMPRESSION_MAX) {
 		if (!(sbi->available_compr_algs & BIT(map->m_algorithmformat))) {
 			erofs_err(inode->i_sb, "inconsistent algorithmtype %u for nid %llu",