Merge tag 'erofs-for-6.4-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/xiang/erofs

Pull erofs updates from Gao Xiang:
 "In this cycle, sub-page block support for uncompressed files is
  available. It's mainly used to enable original signing ('golden')
  4k-block images on arm64 with 16/64k pages. In addition, end users
  could also use this feature to build a manifest to directly refer to
  golden tar data.

  Besides, long xattr name prefix support is also introduced in this
  cycle to avoid too many xattrs with the same prefix (e.g. overlayfs
  xattrs). It's useful for erofs + overlayfs combination (like Composefs
  model): the image size is reduced by ~14% and runtime performance is
  also slightly improved.

  Others are random fixes and cleanups as usual.

  Summary:

   - Add sub-page block size support for uncompressed files

   - Support flattened block device for multi-blob images to be attached
     into virtual machines (including cloud servers) and bare metals

   - Support long xattr name prefixes to optimize images with common
     xattr namespaces (e.g. files with overlayfs xattrs) use cases

   - Various minor cleanups & fixes"

* tag 'erofs-for-6.4-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/xiang/erofs:
  erofs: cleanup i_format-related stuffs
  erofs: sunset erofs_dbg()
  erofs: fix potential overflow calculating xattr_isize
  erofs: get rid of z_erofs_fill_inode()
  erofs: enable long extended attribute name prefixes
  erofs: handle long xattr name prefixes properly
  erofs: add helpers to load long xattr name prefixes
  erofs: introduce on-disk format for long xattr name prefixes
  erofs: move packed inode out of the compression part
  erofs: keep meta inode into erofs_buf
  erofs: initialize packed inode after root inode is assigned
  erofs: stop parsing non-compact HEAD index if clusterofs is invalid
  erofs: don't warn ztailpacking feature anymore
  erofs: simplify erofs_xattr_generic_get()
  erofs: rename init_inode_xattrs with erofs_ prefix
  erofs: move several xattr helpers into xattr.c
  erofs: tidy up EROFS on-disk naming
  erofs: support flattened block device for multi-blob images
  erofs: set block size to the on-disk block size
  erofs: avoid hardcoded blocksize for subpage block support

1 files changed, 161 insertions, 63 deletions

diff --git a/fs/erofs/xattr.c b/fs/erofs/xattr.c
index 015462763bdd..cd80499351e0 100644
--- a/fs/erofs/xattr.c
+++ b/fs/erofs/xattr.c
@@ -7,6 +7,19 @@
 #include <linux/security.h>
 #include "xattr.h"
 
+static inline erofs_blk_t erofs_xattr_blkaddr(struct super_block *sb,
+					      unsigned int xattr_id)
+{
+	return EROFS_SB(sb)->xattr_blkaddr +
+	       erofs_blknr(sb, xattr_id * sizeof(__u32));
+}
+
+static inline unsigned int erofs_xattr_blkoff(struct super_block *sb,
+					      unsigned int xattr_id)
+{
+	return erofs_blkoff(sb, xattr_id * sizeof(__u32));
+}
+
 struct xattr_iter {
 	struct super_block *sb;
 	struct erofs_buf buf;
@@ -16,7 +29,7 @@ struct xattr_iter {
 	unsigned int ofs;
 };
 
-static int init_inode_xattrs(struct inode *inode)
+static int erofs_init_inode_xattrs(struct inode *inode)
 {
 	struct erofs_inode *const vi = EROFS_I(inode);
 	struct xattr_iter it;
@@ -68,8 +81,8 @@ static int init_inode_xattrs(struct inode *inode)
 	}
 
 	it.buf = __EROFS_BUF_INITIALIZER;
-	it.blkaddr = erofs_blknr(erofs_iloc(inode) + vi->inode_isize);
-	it.ofs = erofs_blkoff(erofs_iloc(inode) + vi->inode_isize);
+	it.blkaddr = erofs_blknr(sb, erofs_iloc(inode) + vi->inode_isize);
+	it.ofs = erofs_blkoff(sb, erofs_iloc(inode) + vi->inode_isize);
 
 	/* read in shared xattr array (non-atomic, see kmalloc below) */
 	it.kaddr = erofs_read_metabuf(&it.buf, sb, it.blkaddr, EROFS_KMAP);
@@ -92,9 +105,9 @@ static int init_inode_xattrs(struct inode *inode)
 	it.ofs += sizeof(struct erofs_xattr_ibody_header);
 
 	for (i = 0; i < vi->xattr_shared_count; ++i) {
-		if (it.ofs >= EROFS_BLKSIZ) {
+		if (it.ofs >= sb->s_blocksize) {
 			/* cannot be unaligned */
-			DBG_BUGON(it.ofs != EROFS_BLKSIZ);
+			DBG_BUGON(it.ofs != sb->s_blocksize);
 
 			it.kaddr = erofs_read_metabuf(&it.buf, sb, ++it.blkaddr,
 						      EROFS_KMAP);
@@ -139,15 +152,15 @@ struct xattr_iter_handlers {
 
 static inline int xattr_iter_fixup(struct xattr_iter *it)
 {
-	if (it->ofs < EROFS_BLKSIZ)
+	if (it->ofs < it->sb->s_blocksize)
 		return 0;
 
-	it->blkaddr += erofs_blknr(it->ofs);
+	it->blkaddr += erofs_blknr(it->sb, it->ofs);
 	it->kaddr = erofs_read_metabuf(&it->buf, it->sb, it->blkaddr,
 				       EROFS_KMAP);
 	if (IS_ERR(it->kaddr))
 		return PTR_ERR(it->kaddr);
-	it->ofs = erofs_blkoff(it->ofs);
+	it->ofs = erofs_blkoff(it->sb, it->ofs);
 	return 0;
 }
 
@@ -157,7 +170,8 @@ static int inline_xattr_iter_begin(struct xattr_iter *it,
 	struct erofs_inode *const vi = EROFS_I(inode);
 	unsigned int xattr_header_sz, inline_xattr_ofs;
 
-	xattr_header_sz = inlinexattr_header_size(inode);
+	xattr_header_sz = sizeof(struct erofs_xattr_ibody_header) +
+			  sizeof(u32) * vi->xattr_shared_count;
 	if (xattr_header_sz >= vi->xattr_isize) {
 		DBG_BUGON(xattr_header_sz > vi->xattr_isize);
 		return -ENOATTR;
@@ -165,8 +179,8 @@ static int inline_xattr_iter_begin(struct xattr_iter *it,
 
 	inline_xattr_ofs = vi->inode_isize + xattr_header_sz;
 
-	it->blkaddr = erofs_blknr(erofs_iloc(inode) + inline_xattr_ofs);
-	it->ofs = erofs_blkoff(erofs_iloc(inode) + inline_xattr_ofs);
+	it->blkaddr = erofs_blknr(it->sb, erofs_iloc(inode) + inline_xattr_ofs);
+	it->ofs = erofs_blkoff(it->sb, erofs_iloc(inode) + inline_xattr_ofs);
 	it->kaddr = erofs_read_metabuf(&it->buf, inode->i_sb, it->blkaddr,
 				       EROFS_KMAP);
 	if (IS_ERR(it->kaddr))
@@ -222,8 +236,8 @@ static int xattr_foreach(struct xattr_iter *it,
 	processed = 0;
 
 	while (processed < entry.e_name_len) {
-		if (it->ofs >= EROFS_BLKSIZ) {
-			DBG_BUGON(it->ofs > EROFS_BLKSIZ);
+		if (it->ofs >= it->sb->s_blocksize) {
+			DBG_BUGON(it->ofs > it->sb->s_blocksize);
 
 			err = xattr_iter_fixup(it);
 			if (err)
@@ -231,7 +245,7 @@ static int xattr_foreach(struct xattr_iter *it,
 			it->ofs = 0;
 		}
 
-		slice = min_t(unsigned int, EROFS_BLKSIZ - it->ofs,
+		slice = min_t(unsigned int, it->sb->s_blocksize - it->ofs,
 			      entry.e_name_len - processed);
 
 		/* handle name */
@@ -257,8 +271,8 @@ static int xattr_foreach(struct xattr_iter *it,
 	}
 
 	while (processed < value_sz) {
-		if (it->ofs >= EROFS_BLKSIZ) {
-			DBG_BUGON(it->ofs > EROFS_BLKSIZ);
+		if (it->ofs >= it->sb->s_blocksize) {
+			DBG_BUGON(it->ofs > it->sb->s_blocksize);
 
 			err = xattr_iter_fixup(it);
 			if (err)
@@ -266,7 +280,7 @@ static int xattr_foreach(struct xattr_iter *it,
 			it->ofs = 0;
 		}
 
-		slice = min_t(unsigned int, EROFS_BLKSIZ - it->ofs,
+		slice = min_t(unsigned int, it->sb->s_blocksize - it->ofs,
 			      value_sz - processed);
 		op->value(it, processed, it->kaddr + it->ofs, slice);
 		it->ofs += slice;
@@ -283,17 +297,45 @@ struct getxattr_iter {
 	struct xattr_iter it;
 
 	char *buffer;
-	int buffer_size, index;
+	int buffer_size, index, infix_len;
 	struct qstr name;
 };
 
+static int erofs_xattr_long_entrymatch(struct getxattr_iter *it,
+				       struct erofs_xattr_entry *entry)
+{
+	struct erofs_sb_info *sbi = EROFS_SB(it->it.sb);
+	struct erofs_xattr_prefix_item *pf = sbi->xattr_prefixes +
+		(entry->e_name_index & EROFS_XATTR_LONG_PREFIX_MASK);
+
+	if (pf >= sbi->xattr_prefixes + sbi->xattr_prefix_count)
+		return -ENOATTR;
+
+	if (it->index != pf->prefix->base_index ||
+	    it->name.len != entry->e_name_len + pf->infix_len)
+		return -ENOATTR;
+
+	if (memcmp(it->name.name, pf->prefix->infix, pf->infix_len))
+		return -ENOATTR;
+
+	it->infix_len = pf->infix_len;
+	return 0;
+}
+
 static int xattr_entrymatch(struct xattr_iter *_it,
 			    struct erofs_xattr_entry *entry)
 {
 	struct getxattr_iter *it = container_of(_it, struct getxattr_iter, it);
 
-	return (it->index != entry->e_name_index ||
-		it->name.len != entry->e_name_len) ? -ENOATTR : 0;
+	/* should also match the infix for long name prefixes */
+	if (entry->e_name_index & EROFS_XATTR_LONG_PREFIX)
+		return erofs_xattr_long_entrymatch(it, entry);
+
+	if (it->index != entry->e_name_index ||
+	    it->name.len != entry->e_name_len)
+		return -ENOATTR;
+	it->infix_len = 0;
+	return 0;
 }
 
 static int xattr_namematch(struct xattr_iter *_it,
@@ -301,7 +343,9 @@ static int xattr_namematch(struct xattr_iter *_it,
 {
 	struct getxattr_iter *it = container_of(_it, struct getxattr_iter, it);
 
-	return memcmp(buf, it->name.name + processed, len) ? -ENOATTR : 0;
+	if (memcmp(buf, it->name.name + it->infix_len + processed, len))
+		return -ENOATTR;
+	return 0;
 }
 
 static int xattr_checkbuffer(struct xattr_iter *_it,
@@ -351,21 +395,18 @@ static int inline_getxattr(struct inode *inode, struct getxattr_iter *it)
 static int shared_getxattr(struct inode *inode, struct getxattr_iter *it)
 {
 	struct erofs_inode *const vi = EROFS_I(inode);
-	struct super_block *const sb = inode->i_sb;
-	struct erofs_sb_info *const sbi = EROFS_SB(sb);
-	unsigned int i;
+	struct super_block *const sb = it->it.sb;
+	unsigned int i, xsid;
 	int ret = -ENOATTR;
 
 	for (i = 0; i < vi->xattr_shared_count; ++i) {
-		erofs_blk_t blkaddr =
-			xattrblock_addr(sbi, vi->xattr_shared_xattrs[i]);
-
-		it->it.ofs = xattrblock_offset(sbi, vi->xattr_shared_xattrs[i]);
-		it->it.kaddr = erofs_read_metabuf(&it->it.buf, sb, blkaddr,
-						  EROFS_KMAP);
+		xsid = vi->xattr_shared_xattrs[i];
+		it->it.blkaddr = erofs_xattr_blkaddr(sb, xsid);
+		it->it.ofs = erofs_xattr_blkoff(sb, xsid);
+		it->it.kaddr = erofs_read_metabuf(&it->it.buf, sb,
+						  it->it.blkaddr, EROFS_KMAP);
 		if (IS_ERR(it->it.kaddr))
 			return PTR_ERR(it->it.kaddr);
-		it->it.blkaddr = blkaddr;
 
 		ret = xattr_foreach(&it->it, &find_xattr_handlers, NULL);
 		if (ret != -ENOATTR)
@@ -394,7 +435,7 @@ int erofs_getxattr(struct inode *inode, int index,
 	if (!name)
 		return -EINVAL;
 
-	ret = init_inode_xattrs(inode);
+	ret = erofs_init_inode_xattrs(inode);
 	if (ret)
 		return ret;
 
@@ -421,20 +462,9 @@ static int erofs_xattr_generic_get(const struct xattr_handler *handler,
 				   struct dentry *unused, struct inode *inode,
 				   const char *name, void *buffer, size_t size)
 {
-	struct erofs_sb_info *const sbi = EROFS_I_SB(inode);
-
-	switch (handler->flags) {
-	case EROFS_XATTR_INDEX_USER:
-		if (!test_opt(&sbi->opt, XATTR_USER))
-			return -EOPNOTSUPP;
-		break;
-	case EROFS_XATTR_INDEX_TRUSTED:
-		break;
-	case EROFS_XATTR_INDEX_SECURITY:
-		break;
-	default:
-		return -EINVAL;
-	}
+	if (handler->flags == EROFS_XATTR_INDEX_USER &&
+	    !test_opt(&EROFS_I_SB(inode)->opt, XATTR_USER))
+		return -EOPNOTSUPP;
 
 	return erofs_getxattr(inode, handler->flags, name, buffer, size);
 }
@@ -483,25 +513,40 @@ static int xattr_entrylist(struct xattr_iter *_it,
 {
 	struct listxattr_iter *it =
 		container_of(_it, struct listxattr_iter, it);
-	unsigned int prefix_len;
-	const char *prefix;
+	unsigned int base_index = entry->e_name_index;
+	unsigned int prefix_len, infix_len = 0;
+	const char *prefix, *infix = NULL;
+
+	if (entry->e_name_index & EROFS_XATTR_LONG_PREFIX) {
+		struct erofs_sb_info *sbi = EROFS_SB(_it->sb);
+		struct erofs_xattr_prefix_item *pf = sbi->xattr_prefixes +
+			(entry->e_name_index & EROFS_XATTR_LONG_PREFIX_MASK);
+
+		if (pf >= sbi->xattr_prefixes + sbi->xattr_prefix_count)
+			return 1;
+		infix = pf->prefix->infix;
+		infix_len = pf->infix_len;
+		base_index = pf->prefix->base_index;
+	}
 
-	prefix = erofs_xattr_prefix(entry->e_name_index, it->dentry);
+	prefix = erofs_xattr_prefix(base_index, it->dentry);
 	if (!prefix)
 		return 1;
 	prefix_len = strlen(prefix);
 
 	if (!it->buffer) {
-		it->buffer_ofs += prefix_len + entry->e_name_len + 1;
+		it->buffer_ofs += prefix_len + infix_len +
+					entry->e_name_len + 1;
 		return 1;
 	}
 
-	if (it->buffer_ofs + prefix_len
+	if (it->buffer_ofs + prefix_len + infix_len +
 		+ entry->e_name_len + 1 > it->buffer_size)
 		return -ERANGE;
 
 	memcpy(it->buffer + it->buffer_ofs, prefix, prefix_len);
-	it->buffer_ofs += prefix_len;
+	memcpy(it->buffer + it->buffer_ofs + prefix_len, infix, infix_len);
+	it->buffer_ofs += prefix_len + infix_len;
 	return 0;
 }
 
@@ -555,21 +600,18 @@ static int shared_listxattr(struct listxattr_iter *it)
 {
 	struct inode *const inode = d_inode(it->dentry);
 	struct erofs_inode *const vi = EROFS_I(inode);
-	struct super_block *const sb = inode->i_sb;
-	struct erofs_sb_info *const sbi = EROFS_SB(sb);
-	unsigned int i;
+	struct super_block *const sb = it->it.sb;
+	unsigned int i, xsid;
 	int ret = 0;
 
 	for (i = 0; i < vi->xattr_shared_count; ++i) {
-		erofs_blk_t blkaddr =
-			xattrblock_addr(sbi, vi->xattr_shared_xattrs[i]);
-
-		it->it.ofs = xattrblock_offset(sbi, vi->xattr_shared_xattrs[i]);
-		it->it.kaddr = erofs_read_metabuf(&it->it.buf, sb, blkaddr,
-						  EROFS_KMAP);
+		xsid = vi->xattr_shared_xattrs[i];
+		it->it.blkaddr = erofs_xattr_blkaddr(sb, xsid);
+		it->it.ofs = erofs_xattr_blkoff(sb, xsid);
+		it->it.kaddr = erofs_read_metabuf(&it->it.buf, sb,
+						  it->it.blkaddr, EROFS_KMAP);
 		if (IS_ERR(it->it.kaddr))
 			return PTR_ERR(it->it.kaddr);
-		it->it.blkaddr = blkaddr;
 
 		ret = xattr_foreach(&it->it, &list_xattr_handlers, NULL);
 		if (ret)
@@ -584,7 +626,7 @@ ssize_t erofs_listxattr(struct dentry *dentry,
 	int ret;
 	struct listxattr_iter it;
 
-	ret = init_inode_xattrs(d_inode(dentry));
+	ret = erofs_init_inode_xattrs(d_inode(dentry));
 	if (ret == -ENOATTR)
 		return 0;
 	if (ret)
@@ -605,6 +647,62 @@ ssize_t erofs_listxattr(struct dentry *dentry,
 	return ret;
 }
 
+void erofs_xattr_prefixes_cleanup(struct super_block *sb)
+{
+	struct erofs_sb_info *sbi = EROFS_SB(sb);
+	int i;
+
+	if (sbi->xattr_prefixes) {
+		for (i = 0; i < sbi->xattr_prefix_count; i++)
+			kfree(sbi->xattr_prefixes[i].prefix);
+		kfree(sbi->xattr_prefixes);
+		sbi->xattr_prefixes = NULL;
+	}
+}
+
+int erofs_xattr_prefixes_init(struct super_block *sb)
+{
+	struct erofs_sb_info *sbi = EROFS_SB(sb);
+	struct erofs_buf buf = __EROFS_BUF_INITIALIZER;
+	erofs_off_t pos = (erofs_off_t)sbi->xattr_prefix_start << 2;
+	struct erofs_xattr_prefix_item *pfs;
+	int ret = 0, i, len;
+
+	if (!sbi->xattr_prefix_count)
+		return 0;
+
+	pfs = kzalloc(sbi->xattr_prefix_count * sizeof(*pfs), GFP_KERNEL);
+	if (!pfs)
+		return -ENOMEM;
+
+	if (erofs_sb_has_fragments(sbi))
+		buf.inode = sbi->packed_inode;
+	else
+		erofs_init_metabuf(&buf, sb);
+
+	for (i = 0; i < sbi->xattr_prefix_count; i++) {
+		void *ptr = erofs_read_metadata(sb, &buf, &pos, &len);
+
+		if (IS_ERR(ptr)) {
+			ret = PTR_ERR(ptr);
+			break;
+		} else if (len < sizeof(*pfs->prefix) ||
+			   len > EROFS_NAME_LEN + sizeof(*pfs->prefix)) {
+			kfree(ptr);
+			ret = -EFSCORRUPTED;
+			break;
+		}
+		pfs[i].prefix = ptr;
+		pfs[i].infix_len = len - sizeof(struct erofs_xattr_long_prefix);
+	}
+
+	erofs_put_metabuf(&buf);
+	sbi->xattr_prefixes = pfs;
+	if (ret)
+		erofs_xattr_prefixes_cleanup(sb);
+	return ret;
+}
+
 #ifdef CONFIG_EROFS_FS_POSIX_ACL
 struct posix_acl *erofs_get_acl(struct inode *inode, int type, bool rcu)
 {